https://korshunov.ai/en/hourly1.0 https://korshunov.ai/ru/hourly1.0 https://korshunov.ai/en/today/hourly0.9 https://korshunov.ai/en/week/daily0.9 https://korshunov.ai/en/all/daily0.8 https://korshunov.ai/ru/today/hourly0.9 https://korshunov.ai/ru/week/daily0.9 https://korshunov.ai/ru/all/daily0.8 https://korshunov.ai/en/lab/01-ai/daily0.8 https://korshunov.ai/en/lab/ai2/daily0.8 https://korshunov.ai/en/lab/alibaba-qwen/daily0.8 https://korshunov.ai/en/lab/anthropic/daily0.8 https://korshunov.ai/en/lab/apple-mlx/daily0.8 https://korshunov.ai/en/lab/baidu/daily0.8 https://korshunov.ai/en/lab/bytedance/daily0.8 https://korshunov.ai/en/lab/cognition/daily0.8 https://korshunov.ai/en/lab/cohere/daily0.8 https://korshunov.ai/en/lab/cursor/daily0.8 https://korshunov.ai/en/lab/deepseek/daily0.8 https://korshunov.ai/en/lab/google-deepmind/daily0.8 https://korshunov.ai/en/lab/hugging-face/daily0.8 https://korshunov.ai/en/lab/meta-ai/daily0.8 https://korshunov.ai/en/lab/microsoft-research/daily0.8 https://korshunov.ai/en/lab/mistral/daily0.8 https://korshunov.ai/en/lab/moonshot/daily0.8 https://korshunov.ai/en/lab/nous/daily0.8 https://korshunov.ai/en/lab/nvidia/daily0.8 https://korshunov.ai/en/lab/openai/daily0.8 https://korshunov.ai/en/lab/perplexity/daily0.8 https://korshunov.ai/en/lab/sakana/daily0.8 https://korshunov.ai/en/lab/stability/daily0.8 https://korshunov.ai/en/lab/xai/daily0.8 https://korshunov.ai/en/lab/zhipu/daily0.8 https://korshunov.ai/en/topic/agents/daily0.8 https://korshunov.ai/en/topic/api-product/daily0.8 https://korshunov.ai/en/topic/benchmarks-results/daily0.8 https://korshunov.ai/en/topic/coding/daily0.8 https://korshunov.ai/en/topic/data/daily0.8 https://korshunov.ai/en/topic/evaluation/daily0.8 https://korshunov.ai/en/topic/hardware/daily0.8 https://korshunov.ai/en/topic/image/daily0.8 https://korshunov.ai/en/topic/inference-efficiency/daily0.8 https://korshunov.ai/en/topic/multimodal/daily0.8 https://korshunov.ai/en/topic/open-weights/daily0.8 https://korshunov.ai/en/topic/policy-regulation/daily0.8 https://korshunov.ai/en/topic/rag/daily0.8 https://korshunov.ai/en/topic/reasoning/daily0.8 https://korshunov.ai/en/topic/research-paper/daily0.8 https://korshunov.ai/en/topic/robotics/daily0.8 https://korshunov.ai/en/topic/safety-alignment/daily0.8 https://korshunov.ai/en/topic/training-methods/daily0.8 https://korshunov.ai/en/topic/video/daily0.8 https://korshunov.ai/en/topic/voice-audio/daily0.8 https://korshunov.ai/en/source/anthropic-news/daily0.7 https://korshunov.ai/en/source/arxiv-cs-ai/daily0.7 https://korshunov.ai/en/source/arxiv-cs-cl/daily0.7 https://korshunov.ai/en/source/arxiv-cs-lg/daily0.7 https://korshunov.ai/en/source/autogpt/daily0.7 https://korshunov.ai/en/source/claude-code-releases/daily0.7 https://korshunov.ai/en/source/cohere-blog/daily0.7 https://korshunov.ai/en/source/crewai/daily0.7 https://korshunov.ai/en/source/deepmind-blog/daily0.7 https://korshunov.ai/en/source/dwatv/daily0.7 https://korshunov.ai/en/source/eugene-yan/daily0.7 https://korshunov.ai/en/source/google-ai-keyword/daily0.7 https://korshunov.ai/en/source/goose/daily0.7 https://korshunov.ai/en/source/hf-forums/daily0.7 https://korshunov.ai/en/source/huggingface-blog/daily0.7 https://korshunov.ai/en/source/import-ai/daily0.7 https://korshunov.ai/en/source/interconnects/daily0.7 https://korshunov.ai/en/source/langgraph/daily0.7 https://korshunov.ai/en/source/latent-space/daily0.7 https://korshunov.ai/en/source/llama-cpp/daily0.7 https://korshunov.ai/en/source/llamaindex/daily0.7 https://korshunov.ai/en/source/marktechpost/daily0.7 https://korshunov.ai/en/source/microsoft-research-blog/daily0.7 https://korshunov.ai/en/source/mistral-news/daily0.7 https://korshunov.ai/en/source/nvidia-tech-blog/daily0.7 https://korshunov.ai/en/source/open-interpreter/daily0.7 https://korshunov.ai/en/source/openai-agents-sdk/daily0.7 https://korshunov.ai/en/source/openai-news/daily0.7 https://korshunov.ai/en/source/r-localllama/daily0.7 https://korshunov.ai/en/source/simon-willison/daily0.7 https://korshunov.ai/en/source/smol-ai-news/daily0.7 https://korshunov.ai/en/source/tldr-ai/daily0.7 https://korshunov.ai/en/source/together/daily0.7 https://korshunov.ai/en/source/vllm/daily0.7 https://korshunov.ai/en/benchmark/agentbench/daily0.8 https://korshunov.ai/en/benchmark/aime-2024/daily0.8 https://korshunov.ai/en/benchmark/aime-2025/daily0.8 https://korshunov.ai/en/benchmark/arc-agi-1/daily0.8 https://korshunov.ai/en/benchmark/bbh/daily0.8 https://korshunov.ai/en/benchmark/browsecomp/daily0.8 https://korshunov.ai/en/benchmark/crux-eval/daily0.8 https://korshunov.ai/en/benchmark/gaia/daily0.8 https://korshunov.ai/en/benchmark/gpqa-diamond/daily0.8 https://korshunov.ai/en/benchmark/gsm8k/daily0.8 https://korshunov.ai/en/benchmark/harmbench/daily0.8 https://korshunov.ai/en/benchmark/humaneval/daily0.8 https://korshunov.ai/en/benchmark/humaneval-plus/daily0.8 https://korshunov.ai/en/benchmark/ifeval/daily0.8 https://korshunov.ai/en/benchmark/livecodebench/daily0.8 https://korshunov.ai/en/benchmark/lmsys-arena/daily0.8 https://korshunov.ai/en/benchmark/longvideobench/daily0.8 https://korshunov.ai/en/benchmark/math-500/daily0.8 https://korshunov.ai/en/benchmark/mbpp/daily0.8 https://korshunov.ai/en/benchmark/mbpp-plus/daily0.8 https://korshunov.ai/en/benchmark/mmlu-pro/daily0.8 https://korshunov.ai/en/benchmark/multi-swe-bench/daily0.8 https://korshunov.ai/en/benchmark/osworld/daily0.8 https://korshunov.ai/en/benchmark/putnambench/daily0.8 https://korshunov.ai/en/benchmark/swe-bench/daily0.8 https://korshunov.ai/en/benchmark/swe-bench-multilingual/daily0.8 https://korshunov.ai/en/benchmark/swe-bench-verified/daily0.8 https://korshunov.ai/en/benchmark/tau-bench/daily0.8 https://korshunov.ai/en/benchmark/tau2-bench/daily0.8 https://korshunov.ai/en/benchmark/terminal-bench/daily0.8 https://korshunov.ai/en/benchmark/webarena/daily0.8 https://korshunov.ai/ru/lab/01-ai/daily0.8 https://korshunov.ai/ru/lab/ai2/daily0.8 https://korshunov.ai/ru/lab/alibaba-qwen/daily0.8 https://korshunov.ai/ru/lab/anthropic/daily0.8 https://korshunov.ai/ru/lab/apple-mlx/daily0.8 https://korshunov.ai/ru/lab/baidu/daily0.8 https://korshunov.ai/ru/lab/bytedance/daily0.8 https://korshunov.ai/ru/lab/cognition/daily0.8 https://korshunov.ai/ru/lab/cohere/daily0.8 https://korshunov.ai/ru/lab/cursor/daily0.8 https://korshunov.ai/ru/lab/deepseek/daily0.8 https://korshunov.ai/ru/lab/google-deepmind/daily0.8 https://korshunov.ai/ru/lab/hugging-face/daily0.8 https://korshunov.ai/ru/lab/meta-ai/daily0.8 https://korshunov.ai/ru/lab/microsoft-research/daily0.8 https://korshunov.ai/ru/lab/mistral/daily0.8 https://korshunov.ai/ru/lab/moonshot/daily0.8 https://korshunov.ai/ru/lab/nous/daily0.8 https://korshunov.ai/ru/lab/nvidia/daily0.8 https://korshunov.ai/ru/lab/openai/daily0.8 https://korshunov.ai/ru/lab/perplexity/daily0.8 https://korshunov.ai/ru/lab/sakana/daily0.8 https://korshunov.ai/ru/lab/stability/daily0.8 https://korshunov.ai/ru/lab/xai/daily0.8 https://korshunov.ai/ru/lab/zhipu/daily0.8 https://korshunov.ai/ru/topic/agents/daily0.8 https://korshunov.ai/ru/topic/api-product/daily0.8 https://korshunov.ai/ru/topic/benchmarks-results/daily0.8 https://korshunov.ai/ru/topic/coding/daily0.8 https://korshunov.ai/ru/topic/data/daily0.8 https://korshunov.ai/ru/topic/evaluation/daily0.8 https://korshunov.ai/ru/topic/hardware/daily0.8 https://korshunov.ai/ru/topic/image/daily0.8 https://korshunov.ai/ru/topic/inference-efficiency/daily0.8 https://korshunov.ai/ru/topic/multimodal/daily0.8 https://korshunov.ai/ru/topic/open-weights/daily0.8 https://korshunov.ai/ru/topic/policy-regulation/daily0.8 https://korshunov.ai/ru/topic/rag/daily0.8 https://korshunov.ai/ru/topic/reasoning/daily0.8 https://korshunov.ai/ru/topic/research-paper/daily0.8 https://korshunov.ai/ru/topic/robotics/daily0.8 https://korshunov.ai/ru/topic/safety-alignment/daily0.8 https://korshunov.ai/ru/topic/training-methods/daily0.8 https://korshunov.ai/ru/topic/video/daily0.8 https://korshunov.ai/ru/topic/voice-audio/daily0.8 https://korshunov.ai/ru/source/anthropic-news/daily0.7 https://korshunov.ai/ru/source/arxiv-cs-ai/daily0.7 https://korshunov.ai/ru/source/arxiv-cs-cl/daily0.7 https://korshunov.ai/ru/source/arxiv-cs-lg/daily0.7 https://korshunov.ai/ru/source/autogpt/daily0.7 https://korshunov.ai/ru/source/claude-code-releases/daily0.7 https://korshunov.ai/ru/source/cohere-blog/daily0.7 https://korshunov.ai/ru/source/crewai/daily0.7 https://korshunov.ai/ru/source/deepmind-blog/daily0.7 https://korshunov.ai/ru/source/dwatv/daily0.7 https://korshunov.ai/ru/source/eugene-yan/daily0.7 https://korshunov.ai/ru/source/google-ai-keyword/daily0.7 https://korshunov.ai/ru/source/goose/daily0.7 https://korshunov.ai/ru/source/hf-forums/daily0.7 https://korshunov.ai/ru/source/huggingface-blog/daily0.7 https://korshunov.ai/ru/source/import-ai/daily0.7 https://korshunov.ai/ru/source/interconnects/daily0.7 https://korshunov.ai/ru/source/langgraph/daily0.7 https://korshunov.ai/ru/source/latent-space/daily0.7 https://korshunov.ai/ru/source/llama-cpp/daily0.7 https://korshunov.ai/ru/source/llamaindex/daily0.7 https://korshunov.ai/ru/source/marktechpost/daily0.7 https://korshunov.ai/ru/source/microsoft-research-blog/daily0.7 https://korshunov.ai/ru/source/mistral-news/daily0.7 https://korshunov.ai/ru/source/nvidia-tech-blog/daily0.7 https://korshunov.ai/ru/source/open-interpreter/daily0.7 https://korshunov.ai/ru/source/openai-agents-sdk/daily0.7 https://korshunov.ai/ru/source/openai-news/daily0.7 https://korshunov.ai/ru/source/r-localllama/daily0.7 https://korshunov.ai/ru/source/simon-willison/daily0.7 https://korshunov.ai/ru/source/smol-ai-news/daily0.7 https://korshunov.ai/ru/source/tldr-ai/daily0.7 https://korshunov.ai/ru/source/together/daily0.7 https://korshunov.ai/ru/source/vllm/daily0.7 https://korshunov.ai/ru/benchmark/agentbench/daily0.8 https://korshunov.ai/ru/benchmark/aime-2024/daily0.8 https://korshunov.ai/ru/benchmark/aime-2025/daily0.8 https://korshunov.ai/ru/benchmark/arc-agi-1/daily0.8 https://korshunov.ai/ru/benchmark/bbh/daily0.8 https://korshunov.ai/ru/benchmark/browsecomp/daily0.8 https://korshunov.ai/ru/benchmark/crux-eval/daily0.8 https://korshunov.ai/ru/benchmark/gaia/daily0.8 https://korshunov.ai/ru/benchmark/gpqa-diamond/daily0.8 https://korshunov.ai/ru/benchmark/gsm8k/daily0.8 https://korshunov.ai/ru/benchmark/harmbench/daily0.8 https://korshunov.ai/ru/benchmark/humaneval/daily0.8 https://korshunov.ai/ru/benchmark/humaneval-plus/daily0.8 https://korshunov.ai/ru/benchmark/ifeval/daily0.8 https://korshunov.ai/ru/benchmark/livecodebench/daily0.8 https://korshunov.ai/ru/benchmark/lmsys-arena/daily0.8 https://korshunov.ai/ru/benchmark/longvideobench/daily0.8 https://korshunov.ai/ru/benchmark/math-500/daily0.8 https://korshunov.ai/ru/benchmark/mbpp/daily0.8 https://korshunov.ai/ru/benchmark/mbpp-plus/daily0.8 https://korshunov.ai/ru/benchmark/mmlu-pro/daily0.8 https://korshunov.ai/ru/benchmark/multi-swe-bench/daily0.8 https://korshunov.ai/ru/benchmark/osworld/daily0.8 https://korshunov.ai/ru/benchmark/putnambench/daily0.8 https://korshunov.ai/ru/benchmark/swe-bench/daily0.8 https://korshunov.ai/ru/benchmark/swe-bench-multilingual/daily0.8 https://korshunov.ai/ru/benchmark/swe-bench-verified/daily0.8 https://korshunov.ai/ru/benchmark/tau-bench/daily0.8 https://korshunov.ai/ru/benchmark/tau2-bench/daily0.8 https://korshunov.ai/ru/benchmark/terminal-bench/daily0.8 https://korshunov.ai/ru/benchmark/webarena/daily0.8 https://korshunov.ai/en/article/7445-reddit-inquiry-are-third-party-memory-systems-better-than-openclaw-s-built-in/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7445-vopros-na-reddit-luchshe-li-storonnie-sistemy-pamiati-chem-vstroennaia-memory-v/2026-06-25weekly0.7 https://korshunov.ai/en/article/6514-self-filtering-iterative-data-selection-for-vision-language-models/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6514-self-filtering-iterativnyi-otbor-dannykh-dlia-modelei-vizualnyi-iazyk/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7444-apple-povyshaet-tseny-na-vsiu-lineiku-produktov-udvaivaia-stoimost-apgreida/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6512-dit-reward-ispolzovanie-predstavlenii-diffuzionnogo-transformatora-dlia-v-po/2026-06-25weekly0.7 https://korshunov.ai/en/article/6512-dit-reward-using-diffusion-transformer-representations-for-text-to-image-reward/2026-06-25weekly0.7 https://korshunov.ai/en/article/7444-apple-raises-prices-across-product-line-doubling-memory-upgrade-costs/2026-06-25weekly0.7 https://korshunov.ai/en/article/6511-qor-compact-a-five-item-daily-survey-for-remote-patient-monitoring/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6510-indeksy-podverzhennosti-ii-ogranicheniia-staticheskikh-metrik-i-neobkhodimost-i/2026-06-25weekly0.7 https://korshunov.ai/en/article/6510-ai-exposure-scores-limitations-of-static-metrics-and-the-need-for-research/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6511-qor-compact-piatipunktovyi-ezhednevnyi-oprosnik-dlia-udalennogo-monitoringa/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6509-obuchenie-protsessnym-nagradam-cherez-sopostavlenie-chastoty-poseshcheniia-dlia/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6508-tailormind-k-generatsii-multimodalnogo-kontenta-s-uchetom-predpochtenii/2026-06-25weekly0.7 https://korshunov.ai/en/article/6508-tailormind-towards-preference-aligned-multimodal-content-generation/2026-06-25weekly0.7 https://korshunov.ai/en/article/6509-learning-process-rewards-via-success-visitation-matching-for-efficient-rl/2026-06-25weekly0.7 https://korshunov.ai/en/article/6507-tapered-language-models-improving-performance-via-depth-aware-capacity/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6506-chellendzh-nvidia-nemotron-strokovoe-sopostavlenie-i-vozvrat-s-otkatom-dlia/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6507-suzhaiushchiesia-iazykovye-modeli-povyshenie-proizvoditelnosti-za-schet-emkosti/2026-06-25weekly0.7 https://korshunov.ai/en/article/6506-nvidia-nemotron-challenge-string-matching-and-backtracking-for-bit-manipulation/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6505-psybridge-gibridnaia-sistema-dlia-mnogomernoi-otsenki-psikhicheskogo-zdorovia/2026-06-25weekly0.7 https://korshunov.ai/en/article/6505-psybridge-a-hybrid-framework-for-multi-dimensional-mental-health-assessment/2026-06-25weekly0.7 https://korshunov.ai/en/article/6504-open-problem-is-adamw-effective-under-heavy-tailed-noise/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6504-otkrytaia-problema-effektiven-li-adamw-pri-tiazhelokhvostom-shume/2026-06-25weekly0.7 https://korshunov.ai/en/article/6502-semantic-browsing-controllable-diversity-for-image-generation/2026-06-25weekly0.7 https://korshunov.ai/en/article/6503-air-adaptive-interleaved-reasoning-with-code-in-mllms/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6502-semanticheskii-prosmotr-upravliaemoe-raznoobrazie-dlia-generatsii-izobrazhenii/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6503-air-adaptivnoe-chereduiushcheesia-rassuzhdenie-s-kodom-v-multimodalnykh-bolshikh/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6501-coordex-koordinatsiia-apriornykh-znanii-o-tele-i-kistiakh-dlia-nepreryvnoi-s-u/2026-06-25weekly0.7 https://korshunov.ai/en/article/6501-coordex-coordinating-body-and-hand-priors-for-continuous-dexterous-humanoid-loco/2026-06-25weekly0.7 https://korshunov.ai/en/article/6489-encoder-decoder-manifold-alignment-for-idempotent-generation/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6489-soglasovanie-mnogoobrazii-kodirovshchika-i-dekodera-dlia-idempotentnoi/2026-06-25weekly0.7 https://korshunov.ai/en/article/6488-manifold-restore-mixing-enhances-protein-representation-learning/2026-06-25weekly0.7 https://korshunov.ai/en/article/6487-entropy-guided-boundary-supervision-for-breast-ultrasound-segmentation/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6487-entropiino-upravliaemoe-granichnoe-obuchenie-dlia-segmentatsii-ultrazvukovykh/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6488-manifold-restore-mixing-uluchshaet-obuchenie-predstavlenii-belkov/2026-06-25weekly0.7 https://korshunov.ai/en/article/6486-diffusion-integrated-gradients-controllable-path-generation-for-flexible-feature/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6483-pervyi-analiz-konechnogo-vremeni-dlia-klassicheskogo-adam-v-negladkoi-nevypukloi/2026-06-25weekly0.7 https://korshunov.ai/en/article/6483-first-finite-time-analysis-of-classical-adam-for-nonsmooth-nonconvex/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6486-diffuzionnye-integrirovannye-gradienty-upravliaemaia-generatsiia-putei-dlia/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6485-kurkuliarnoe-obuchenie-s-podkrepleniem-uchityvaiushchee-granitsy-rasshiriaet-k/2026-06-25weekly0.7 https://korshunov.ai/en/article/6485-boundary-aware-curriculum-rl-expands-llm-reasoning-capacity-beyond-base-model/2026-06-25weekly0.7 https://korshunov.ai/en/article/6484-attention-sinks-and-collapse-are-universal-consequences-of-content-based-routing/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6484-sinki-vnimaniia-i-kollaps-universalnye-sledstviia-marshrutizatsii-na-osnove/2026-06-25weekly0.7 https://korshunov.ai/en/article/7442-inquiry-regarding-the-availability-of-modern-non-chat-completion-models/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7443-polzovateli-soobshchaiut-o-vysokoi-proizvoditelnosti-modeli-siq1-na-kebab-bench/2026-06-25weekly0.7 https://korshunov.ai/en/article/7443-user-reports-strong-performance-of-siq1-model-on-kebab-bench/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7442-zapros-o-nalichii-sovremennykh-modelei-ne-prednaznachennykh-dlia-zaversheniia/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6482-otsutstvie-obobshcheniia-bez-etalona-v-kvantovom-mashinnom-obuchenii/2026-06-25weekly0.7 https://korshunov.ai/en/article/6482-no-reference-free-generalization-in-quantum-machine-learning/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6481-nosimyi-a-rezhimnyi-ultrazvuk-obespechivaet-otslezhivanie-kinematiki-vsei-ruki/2026-06-25weekly0.7 https://korshunov.ai/en/article/6481-wearable-a-mode-ultrasound-enables-whole-hand-kinematic-tracking-on/2026-06-25weekly0.7 https://korshunov.ai/en/article/6480-null-calibrated-conformal-selection-via-target-membership-scores/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6480-nul-kalibrovannyi-konformnyi-otbor-cherez-otsenki-prinadlezhnosti-k-tseli/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6479-robomme-interference-testirovanie-pamiati-robotov-v-usloviiakh-otvlekaiushchikh/2026-06-25weekly0.7 https://korshunov.ai/en/article/6478-flow-annealing-posterior-sampling-for-function-space-regression-and-inverse/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6478-posloinaia-annigiliatsiia-potoka-dlia-semplirovaniia-aposteriornogo-v-regressii/2026-06-25weekly0.7 https://korshunov.ai/en/article/6479-robomme-interference-benchmarks-robot-memory-under-distraction/2026-06-25weekly0.7 https://korshunov.ai/en/article/7435-nvidia-releases-nemotron-twotower-30b-a3b-a-diffusion-based-language-model/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7435-nvidia-vypustila-nemotron-twotower-30b-a3b-iazykovuiu-model-na-osnove-diffuzii/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7441-sempler-i-verifikator-s-vozvratom-nazad-kardinalno-uluchshaiut-proizvoditelnost/2026-06-25weekly0.7 https://korshunov.ai/en/article/7441-backtrack-sampler-and-verifier-drastically-improve-tiny-model-coding-performance/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7434-eksperimentalnaia-realizatsiia-rdma-cherez-usb4-prodemonstrirovana-na-baze-strix/2026-06-25weekly0.7 https://korshunov.ai/en/article/7433-glm-5-2-on-dual-strix-halo-256gb-worth-it/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7433-glm-5-2-na-konfiguratsii-s-dvumia-strix-halo-256-gb-stoit-li-togo/2026-06-25weekly0.7 https://korshunov.ai/en/article/7434-experimental-usb4-rdma-implementation-demonstrated-on-strix-halo/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7432-vopros-na-reddit-o-ispolzovanii-lokalnykh-modelei-dlia-samostoiatelnogo-vzloma/2026-06-25weekly0.7 https://korshunov.ai/en/article/7432-reddit-inquiry-on-using-local-models-for-self-hacking/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7431-polzovateli-soobshchaiut-o-snizhenii-kachestva-i-effektivnosti-modelei-mtp-v-3-6/2026-06-25weekly0.7 https://korshunov.ai/en/article/7431-user-reports-inferior-quality-and-efficiency-with-mtp-models-in-qwen-3-6-and-4/2026-06-25weekly0.7 https://korshunov.ai/en/article/7428-question-on-why-rocm-and-intel-stacks-lag-behind-cuda-in-software-ecosystem/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7428-vopros-o-tom-pochemu-steki-rocm-i-intel-otstaiut-ot-cuda-v-zrelosti-programmnoi/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7430-razrabotchik-prosit-protestirovat-podderzhku-mtp-dlia-glm-4-7-flash-cherez-llama/2026-06-25weekly0.7 https://korshunov.ai/en/article/7430-developer-requests-testing-for-mtp-support-in-glm-4-7-flash-via-llama-cpp/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7426-anthropic-obviniaet-alibaba-v-nezakonnoi-kampanii-po-izvlecheniiu-vozmozhnostei/2026-06-25weekly0.7 https://korshunov.ai/en/article/7426-anthropic-accuses-alibaba-of-illicit-ai-capability-extraction-campaign/2026-06-25weekly0.7 https://korshunov.ai/en/article/7427-community-discussion-on-running-deepseek-v4-flash-with-moe-offload/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7427-obsuzhdenie-v-soobshchestve-po-zapusku-deepseek-v4-flash-s-vygruzkoi-moe/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7425-supraweather-nano-preview-nebolshoi-ft-transformer-dlia-klassifikatsii-pogody/2026-06-25weekly0.7 https://korshunov.ai/en/article/7425-supraweather-nano-preview-a-small-ft-transformer-for-weather-classification/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7369-hipe-2026-izvlechenie-otnoshenii-persona-mesto-iz-mnogoiazychnykh-istoricheskikh/2026-06-25weekly0.7 https://korshunov.ai/en/article/7369-hipe-2026-person-place-relation-extraction-from-multilingual-historical-texts/2026-06-25weekly0.7 https://korshunov.ai/en/article/7368-weave-of-formal-thought-uniting-rigorous-syntactic-validation-with-learned/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7368-tkan-formalnogo-myshleniia-obedinenie-strogogo-sintaksicheskogo-validirovaniia-s/2026-06-25weekly0.7 https://korshunov.ai/en/article/7440-llama-cpp-b9788-adds-sycl-tensor-parallelism-for-dual-gpu-setups/2026-06-25weekly0.7 https://korshunov.ai/en/article/7439-llama-cpp-b9789-release-fixes-moe-quantization-and-provides-multi-platform/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7439-vypusk-llama-cpp-b9789-ispravliaet-kvantovanie-moe-i-predostavliaet-binarnye/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7440-llama-cpp-b9788-dobavliaet-parallelizm-tenzorov-po-sycl-dlia-konfiguratsii-s-gpu/2026-06-25weekly0.7 https://korshunov.ai/en/article/7366-autodata-an-agentic-data-scientist-to-create-high-quality-synthetic-data/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7367-speecheq-otsenka-emotsionalnogo-intellekta-v-sotsialno-orientirovannykh/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7366-autodata-agentnyi-data-saientist-dlia-sozdaniia-vysokokachestvennykh-dannykh/2026-06-25weekly0.7 https://korshunov.ai/en/article/7367-speecheq-benchmarking-emotional-intelligence-in-socially-aware-voice-models/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7438-issledovanie-openai-ii-agenty-transformiruiut-rabotu/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7365-dziri-voicebot-sistema-preobrazovaniia-rechi-v-rech-ot-nachala-do-kontsa-dlia/2026-06-25weekly0.7 https://korshunov.ai/en/article/7438-openai-research-shows-ai-agents-transforming-work/2026-06-25weekly0.7 https://korshunov.ai/en/article/7365-dziri-voicebot-end-to-end-speech-to-speech-system-for-algerian-dialect/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7364-tatoxa-novaia-sistema-detoksikatsii-tekstov-dlia-maloresursnogo-tatarskogo/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7363-kollaps-mnogoshagovogo-ispolzovaniia-instrumentov-v-rl-i-nadzornye-ispravleniia/2026-06-25weekly0.7 https://korshunov.ai/en/article/7363-multi-step-tool-use-rl-collapse-and-supervisory-fixes/2026-06-25weekly0.7 https://korshunov.ai/en/article/7364-tatoxa-a-novel-text-detoxification-system-for-low-resource-tatar/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7361-issledovanie-vyiavilo-chitateli-predpochitaiut-chelovecheskie-literaturnye-na/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7362-obnaruzhenie-otmena-obucheniia-i-vosstanovlenie-zashchita-modelei-summarizatsii/2026-06-25weekly0.7 https://korshunov.ai/en/article/7361-study-finds-readers-prefer-human-over-ai-literary-translations-despite-adequate/2026-06-25weekly0.7 https://korshunov.ai/en/article/7362-detect-unlearn-restore-defending-text-summarization-models-against-data/2026-06-25weekly0.7 https://korshunov.ai/en/article/7436-bro77xp-releases-beginner-friendly-local-ai-vtuber-with-zero-shot-voice-cloning/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7436-bro77xp-vypustil-prostoi-dlia-nachinaiushchikh-lokalnyi-ai-vtuber-s-golosa-v/2026-06-25weekly0.7 https://korshunov.ai/en/article/7360-evaluating-ocr-reasoning-robustness-of-vision-language-models-under-visual/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7360-otsenka-ustoichivosti-k-vizualnym-vozmushcheniiam-ocr-rassuzhdenii-v-modeliakh/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7359-estestvennoe-zabyvanie-asimmetrichnyi-kontrol-togo-kakie-pravila-perezhivaiut/2026-06-25weekly0.7 https://korshunov.ai/en/article/7358-keyword-lexicon-blindness-distorts-rhetorical-stance-measurement/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7358-slepota-k-leksikonu-kliuchevykh-slov-iskazhaet-izmerenie-ritoricheskoi-pozitsii/2026-06-25weekly0.7 https://korshunov.ai/en/article/7359-natural-ungrokking-asymmetric-control-of-which-rules-survive-pretraining/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7357-audit-chuvstvitelnosti-k-poriadku-v-multimodalnykh-bolshikh-iazykovykh-modeliakh/2026-06-25weekly0.7 https://korshunov.ai/en/article/7356-real-time-voice-ai-hears-but-does-not-listen/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7356-golosovoi-ii-v-realnom-vremeni-slyshit-no-ne-slushaet/2026-06-25weekly0.7 https://korshunov.ai/en/article/7357-auditing-order-sensitivity-in-multimodal-large-language-models/2026-06-25weekly0.7 https://korshunov.ai/en/article/7370-locked-dell-quote-for-6x-rtx-pro-6000-max-q-at-8960/2026-06-25weekly0.7 https://korshunov.ai/en/article/7371-local-nl-to-sql-pipeline-using-qwen3-4b-and-deterministic-planning/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7370-zablokirovannoe-kommercheskoe-predlozhenie-dell-na-6x-rtx-pro-6000-max-q-za-8960/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7371-lokalnyi-paiplain-nl-to-sql-s-ispolzovaniem-qwen3-4b-i-determinirovannogo/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7355-zapros-na-reddit-o-zapuske-bolshikh-modelei-s-4-8-videokartami-rtx-6000-pro/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7248-strukturirovanie-arabsko-angliiskogo-mashinochitaemogo-slovaria-s-ispolzovaniem/2026-06-25weekly0.7 https://korshunov.ai/en/article/7248-structuring-an-arabic-english-machine-readable-dictionary-using-parsing-grammars/2026-06-25weekly0.7 https://korshunov.ai/en/article/7355-reddit-inquiry-on-running-large-models-with-4x-8x-rtx-6000-pros/2026-06-25weekly0.7 https://korshunov.ai/en/article/7247-wbcmor-vqa-a-bilingual-english-urdu-hematology-visual-question-answering/2026-06-25weekly0.7 https://korshunov.ai/en/article/7246-automatic-generation-of-highlights-for-academic-paper-via-prompt-based-learning/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7246-avtomaticheskoe-generirovanie-annotatsii-k-nauchnym-statiam-na-osnove-obucheniia/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7247-wbcmor-vqa-dvuiazychnyi-anglo-urdu-benchmark-dlia-vizualnogo-otveta-na-voprosy/2026-06-25weekly0.7 https://korshunov.ai/en/article/7244-data-driven-evolution-of-library-and-information-science-research-methods-1990/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7244-evoliutsiia-metodov-issledovanii-v-bibliotechnom-dele-i-informatike-na-osnove/2026-06-25weekly0.7 https://korshunov.ai/en/article/7245-measuring-research-difficulty-in-nlp-an-inverted-u-shaped-relationship-with/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7245-izmerenie-slozhnosti-issledovanii-v-nlp-obrashchennaia-u-obraznaia-zavisimost-ot/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7242-hybrid-ir-dvukhputnyi-gibridnyi-poisk-s-iterativnym-rassuzhdeniem-dlia-slozhnykh/2026-06-25weekly0.7 https://korshunov.ai/en/article/7242-hybrid-ir-dual-path-hybrid-retrieval-with-iterative-reasoning-for-complex/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7243-illada-8-milliardnaia-iazykovaia-model-s-maskirovannoi-diffuziei-i-polnostiu/2026-06-25weekly0.7 https://korshunov.ai/en/article/7243-illada-an-8b-masked-diffusion-language-model-with-fully-bidirectional-attention/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7240-pamiat-opredeliaet-rezultat-otsenka-togo-kak-razlichnye-roli-pamiati-formiruiut/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7241-lokalnaia-marshrutizatsiia-vetvei-effektivnoe-obuchaemoe-masshtabirovanie-na/2026-06-25weekly0.7 https://korshunov.ai/en/article/7241-local-branch-routing-efficient-trainable-test-time-scaling-for-language-models/2026-06-25weekly0.7 https://korshunov.ai/en/article/7240-memory-makes-the-difference-evaluating-how-different-memory-roles-shape-agents/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7238-sarashina2-2-tts-reshenie-problemy-polifonii-kandzi-v-iaponskom-sinteze-rechi-i/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7239-neironnyi-mashinnyi-perevod-dlia-nizko-resursnoi-pary-tankhul-angliiskii/2026-06-25weekly0.7 https://korshunov.ai/en/article/7238-sarashina2-2-tts-tackling-kanji-polyphony-in-japanese-speech-generation-via-data/2026-06-25weekly0.7 https://korshunov.ai/en/article/7239-neural-machine-translation-for-low-resource-tangkhul-english/2026-06-25weekly0.7 https://korshunov.ai/en/article/7237-computational-stylometry-of-english-pali-canon-translations-across-pitakas/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7237-vychislitelnaia-stilometriia-perevodov-angliiskogo-paliiskogo-kanona-po-trem/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7236-story-operators-dekompozitsiia-preobrazovaniia-ot-originala-k-sikvelu-v/2026-06-25weekly0.7 https://korshunov.ai/en/article/7236-story-operators-decomposing-the-original-to-sequel-transformation-in-embedding/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7234-predstavliaem-korpusy-hlava-cor-i-hlava-ad-variatsii-chelovecheskikh-otsenok-v-i/2026-06-25weekly0.7 https://korshunov.ai/en/article/7235-survey-of-toxicity-detection-and-mitigation-strategies-for-multilingual-language/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7235-obzor-strategii-obnaruzheniia-i-smiagcheniia-toksichnosti-dlia-mnogoiazychnykh/2026-06-25weekly0.7 https://korshunov.ai/en/article/7234-introducing-corpora-hlava-cor-and-hlava-ad-human-label-variation-in-coreference/2026-06-25weekly0.7 https://korshunov.ai/en/article/7233-agent-authored-world-modeling-aligns-training-with-decision-needs/2026-06-25weekly0.7 https://korshunov.ai/en/article/7232-oscillatts-adaptive-oscillatory-inductive-bias-for-modeling-sharp-prosodic-in/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7233-modelirovanie-mira-sozdannoe-agentom-soglasuet-obuchenie-s-potrebnostiami/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7232-oscillatts-adaptivnaia-ostsilliatornaia-induktivnaia-predvziatost-dlia-rezkikh-v/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7429-ispravlenie-upavshikh-iunit-testov-dlia-conv-3d-v-sycl/2026-06-25weekly0.7 https://korshunov.ai/en/article/7231-evaluating-japanese-dialect-robustness-across-speech-and-text-based-large-models/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7231-otsenka-ustoichivosti-k-iaponskim-dialektam-v-rechevykh-i-tekstovykh-bolshikh/2026-06-25weekly0.7 https://korshunov.ai/en/article/7429-fix-failed-unit-test-cases-for-conv-3d-in-sycl/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7230-policyalign-priamoe-soglasovanie-politik-dlia-obespecheniia-bezopasnosti-modelei/2026-06-25weekly0.7 https://korshunov.ai/en/article/7229-translation-enhanced-speech-encoder-pre-training-improves-speech-llms/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7229-predobuchenie-rechevogo-enkodera-s-usileniem-za-schet-perevoda-uluchshaet-modeli/2026-06-25weekly0.7 https://korshunov.ai/en/article/7230-policyalign-direct-policy-based-safety-alignment-for-large-language-models/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7228-proektirovanie-instrumentalnykh-obviazok-i-postobuchenie-agentov-na-osnove/2026-06-25weekly0.7 https://korshunov.ai/en/article/7228-harness-design-and-post-training-in-llm-agents/2026-06-25weekly0.7 https://korshunov.ai/en/article/7227-reclaim-evaluation-shows-lossy-memory-is-worse-than-no-memory/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7227-otsenka-reclaim-pokazyvaet-chto-poteriaiushchaia-pamiat-khuzhe-chem-otsutstvie/2026-06-25weekly0.7 https://korshunov.ai/en/article/7225-probing-self-supervised-speech-representations-on-mandarin-sub-dialects-via/2026-06-25weekly0.7 https://korshunov.ai/en/article/7226-the-generalization-spectrum-a-chromatographic-approach-to-evaluating-learning/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7225-issledovanie-samonadziratelnykh-rechevykh-predstavlenii-dlia-poddialektov-iazyka/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7226-spektr-obobshcheniia-khromatograficheskii-podkhod-k-otsenke-algoritmov/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7224-polnostiu-differentsiruemoe-neironnoe-forsirovannoe-vyravnivanie-s-pomoshchiu/2026-06-25weekly0.7 https://korshunov.ai/en/article/7224-fully-differentiable-neural-forced-alignment-via-soft-dynamic-programming/2026-06-25weekly0.7 https://korshunov.ai/en/article/7223-fine-tuned-pegasus-achieves-state-of-the-art-performance-on-xl-sum-english/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7223-dorabotannaia-model-pegasus-dostigaet-luchshikh-rezultatov-na-korpuse-xl-sum-na/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7221-kalibrovka-i-ustoichivost-k-sostiazatelnym-atakam-avtomatizirovannoi-otsenki-asr/2026-06-25weekly0.7 https://korshunov.ai/en/article/7222-red-teaming-framework-uncovers-llm-faithfulness-vulnerabilities-via-multi-role/2026-06-25weekly0.7 https://korshunov.ai/en/article/7221-calibration-and-adversarial-robustness-of-automated-asr-scoring/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7222-freimvork-dlia-krasnogo-testirovaniia-vyiavliaet-uiazvimosti-vernosti-bolshikh/2026-06-25weekly0.7 https://korshunov.ai/en/article/7219-behavioral-drivers-of-rating-sentiment-incongruence-in-sri-lankan-tourism/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7219-povedencheskie-faktory-nesootvetstviia-reitingov-i-tonalnosti-v-otzyvakh-o-shri/2026-06-25weekly0.7 https://korshunov.ai/en/article/7220-stc-improves-arabic-customer-service-via-marbert-sentiment-analysis/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7220-stc-uluchshaet-obsluzhivanie-klientov-na-arabskom-iazyke-s-pomoshchiu-analiza/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7218-cliff-tokens-vyiavlenie-triggerov-sboev-na-odnom-tokene-v-matematicheskom/2026-06-25weekly0.7 https://korshunov.ai/en/article/7217-swe-pro-benchmark-reveals-significant-gap-between-llms-and-expert-software/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7217-benchmark-swe-pro-vyiavliaet-znachitelnyi-razryv-mezhdu-llm-i-ekspertnoi/2026-06-25weekly0.7 https://korshunov.ai/en/article/7218-cliff-tokens-identifying-single-token-failure-triggers-in-llm-mathematical/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7216-bezopasnost-i-konfidentsialnost-v-generatsii-s-dopolneniem-izvlecheniia-ugrozy-i/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7215-sfl-mtsc-ispolzovanie-semanticheskoi-kadra-urovnevoi-mnogozadachnoi-dlia-rechi-s/2026-06-25weekly0.7 https://korshunov.ai/en/article/7216-security-and-privacy-in-retrieval-augmented-generation-architectures-threats-and/2026-06-25weekly0.7 https://korshunov.ai/en/article/7215-sfl-mtsc-leveraging-semantic-frame-level-multi-task-self-consistency-for-robust/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7213-riazi-8b-bolshaia-iazykovaia-model-dlia-matematicheskogo-rassuzhdeniia-na-urdu/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7214-bipace-optimizatsiia-politiki-s-uchetom-bisimuliatsii-i-kontrfakticheskaia-dlia/2026-06-25weekly0.7 https://korshunov.ai/en/article/7214-bipace-bisimulation-guided-policy-optimization-with-action-counterfactual-for/2026-06-25weekly0.7 https://korshunov.ai/en/article/7213-riazi-8b-an-urdu-large-language-model-for-mathematical-reasoning/2026-06-25weekly0.7 https://korshunov.ai/en/article/7211-reveriemem-perspective-bounded-memory-for-book-based-role-playing-agents/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7211-reveriemem-ogranichennaia-perspektivoi-pamiat-dlia-rolevykh-agentov-na-osnove/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7212-nalog-na-ogranicheniia-v-llm-s-otkrytym-vesom-podavlenie-vyzova-instrumentov-pri/2026-06-25weekly0.7 https://korshunov.ai/en/article/7212-constraint-tax-in-open-weight-llms-tool-calling-suppression-under-structured/2026-06-25weekly0.7 https://korshunov.ai/en/article/7210-medguards-multi-agent-system-for-reliable-medical-error-detection-and-correction/2026-06-25weekly0.7 https://korshunov.ai/en/article/7424-llama-cpp-b9786-release-adds-opencl-non-contiguous-row-support/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7210-medguards-mnogoagentnaia-sistema-dlia-nadezhnogo-obnaruzheniia-i-ispravleniia/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7424-vypusk-llama-cpp-b9786-dobavliaet-podderzhku-nekongruentnykh-strok-cherez-opencl/2026-06-25weekly0.7 https://korshunov.ai/en/article/7209-framework-evaluates-when-graphrag-and-agentic-rag-are-needed/2026-06-25weekly0.7 https://korshunov.ai/en/article/7208-bitembed-extreme-low-bit-framework-for-llm-based-text-embeddings/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7209-freimvork-otsenivaet-kogda-neobkhodimy-graphrag-i-agentnyi-rag/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7208-bitembed-ekstremalno-nizkorazriadnaia-arkhitektura-dlia-tekstovykh-embeddingov/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7206-ras-izmerenie-bezopasnosti-llm-cherez-soglasovanie-s-otkazom/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7207-trace-legkovesnoe-obnaruzhenie-otravleniia-korpusa-v-rag-s-pomoshchiu-atributsii/2026-06-25weekly0.7 https://korshunov.ai/en/article/7207-trace-lightweight-detection-of-corpus-poisoning-in-rag-via-token-influence/2026-06-25weekly0.7 https://korshunov.ai/en/article/7206-ras-measuring-llm-safety-through-refusal-alignment/2026-06-25weekly0.7 https://korshunov.ai/en/article/7205-opera-aligning-open-ended-reasoning-via-objective-perplexity-based-reinforcement/2026-06-25weekly0.7 https://korshunov.ai/en/article/7202-do-encoders-suffice-a-systematic-comparison-of-encoder-and-decoder-safety-judges/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7202-khvataet-li-tolko-enkoderov-sistematicheskoe-sravnenie-sudei-bezopasnosti-na-i/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7205-opera-soglasovanie-otkrytogo-rassuzhdeniia-posredstvom-obektivnogo-obucheniia-s/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7354-niodoo-lokalnyi-rantaim-dlia-upravleniia-skrytym-sostoianiem-zamorozhennykh-llm/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7353-polzovateli-soobshchaiut-o-nedostupnosti-instrumentov-i-mcp-servera-dlia-step-3/2026-06-25weekly0.7 https://korshunov.ai/en/article/7353-user-reports-tool-and-mcp-server-unavailability-for-step-3-7-flash-on/2026-06-25weekly0.7 https://korshunov.ai/en/article/7354-niodoo-a-local-runtime-for-hidden-state-steering-of-frozen-llms/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7350-vypusk-llama-cpp-b9785-s-usilennoi-proverkoi-flagov-i-binarnymi-failami-dlia/2026-06-25weekly0.7 https://korshunov.ai/en/article/7352-prompt-format-inquiry-for-training-unsloth-phi-3-5-mini-instruct/2026-06-25weekly0.7 https://korshunov.ai/en/article/7350-llama-cpp-b9785-release-with-hardened-caps-check-and-multi-platform-binaries/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7352-vopros-o-formate-prompta-dlia-obucheniia-unsloth-phi-3-5-mini-instruct/2026-06-25weekly0.7 https://korshunov.ai/en/article/7203-space-efficient-language-generation-in-the-limit/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7203-ekonomiia-prostranstva-pri-generatsii-iazyka-v-predele/2026-06-25weekly0.7 https://korshunov.ai/en/article/7204-argus-benchmark-evaluates-uncertainty-quantification-stability-across-vision-and/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7204-benchmark-argus-otsenivaet-stabilnost-kolichestvennoi-otsenki-neopredelennosti-i/2026-06-25weekly0.7 https://korshunov.ai/en/article/7200-toolbench-x-benchmarking-tool-using-agents-under-unreliable-environments/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7200-toolbench-x-otsenka-agentov-ispolzuiushchikh-instrumenty-v-nenadezhnykh-sredakh/2026-06-25weekly0.7 https://korshunov.ai/en/article/7199-sara-unlocking-multilingual-knowledge-in-mixture-of-experts-via-semantically/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7199-sara-raskrytie-mnogoiazychnykh-znanii-v-smesi-ekspertov-cherez-semanticheski/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7201-kak-bolshie-iazykovye-modeli-formiruiut-reputatsiiu-brendov-v-raznykh-iazykakh-i/2026-06-25weekly0.7 https://korshunov.ai/en/article/7201-how-large-language-models-source-brand-reputation-across-languages-and-markets/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6476-safer-adaptivnoe-ansamblirovanie-upravliaemoe-nadezhnostiu-dlia-ustoichivoi-vo/2026-06-25weekly0.7 https://korshunov.ai/en/article/6476-safer-reliability-guided-adaptive-ensembling-for-robust-test-time-adaptation/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6477-select-to-act-ierarkhicheskoe-obuchenie-s-podkrepleniem-cherez-adaptivnoe/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6475-ekonomno-aktiviruemoe-obuchenie-slovariu-sviazyvaet-razrezhennost-i-khranenie-s/2026-06-25weekly0.7 https://korshunov.ai/en/article/6475-parsimoniously-activated-dictionary-learning-links-sparsity-and-storage-to/2026-06-25weekly0.7 https://korshunov.ai/en/article/6477-select-to-act-hierarchical-reinforcement-learning-via-adaptive-language-guidance/2026-06-25weekly0.7 https://korshunov.ai/en/article/6472-kiwano-an-open-source-pytorch-toolkit-for-speaker-verification-research/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6472-kiwano-otkrytaia-biblioteka-pytorch-dlia-issledovanii-v-oblasti-verifikatsii/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6473-otsenka-fizicheskoi-soglasovannosti-v-generatsii-video-na-osnove-mirovykh-bez/2026-06-25weekly0.7 https://korshunov.ai/en/article/6473-reference-free-assessment-of-physical-consistency-in-world-model-based-video/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6474-orbit-obuchenie-bez-doobucheniia-dlia-mnogoatributivnogo-povedencheskogo/2026-06-25weekly0.7 https://korshunov.ai/en/article/6474-orbit-training-free-multi-attribute-behavioral-steering-via-orthogonal-subspace/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7198-opros-soobshchestva-o-zapuske-dwarfstar-s-modeliu-deepseek-v4-flash-na-dgx-spark/2026-06-25weekly0.7 https://korshunov.ai/en/article/6471-multigrid-training-for-molecular-generation-using-graph-neural-networks/2026-06-25weekly0.7 https://korshunov.ai/en/article/7197-gemma4-26b-a4b-31b-qat-uncensored-balanced-released-with-mtp-speed-boosts/2026-06-25weekly0.7 https://korshunov.ai/en/article/7198-community-inquiry-on-running-dwarfstar-with-deepseek-v4-flash-on-dgx-spark/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7197-vypushcheny-neseksurizirovannye-sbalansirovannye-gemma4-26b-a4b-i-31b-qat-so-za/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6471-mnogosetochnoe-obuchenie-dlia-generatsii-molekul-s-ispolzovaniem-grafovykh-setei/2026-06-25weekly0.7 https://korshunov.ai/en/article/6470-hyperadapter-structured-hyperedge-adaptation-for-parameter-efficient-fine-tuning/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6470-hyperadapter-strukturirovannaia-adaptatsiia-giperreber-dlia-parametricheski/2026-06-25weekly0.7 https://korshunov.ai/en/article/6468-efficient-cnn-with-transfer-learning-for-multi-cancer-detection/2026-06-25weekly0.7 https://korshunov.ai/en/article/6469-shift-invariant-variance-estimator-eliminates-minimization-bias-in-local/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6469-smeshchenno-invariantnyi-otsenitel-dispersii-ustraniaet-smenie-minimizatsii-v/2026-06-25weekly0.7 https://korshunov.ai/ru/article/7196-saimon-uillison-preobrazuet-dannye-o-sovmestimosti-brauzerov-mdn-v-bazu-dannykh/2026-06-25weekly0.7 https://korshunov.ai/en/article/7196-simon-willison-converts-mdn-browser-compatibility-data-into-a-sqlite-database/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6468-effektivnyi-cnn-s-transfernym-obucheniem-dlia-obnaruzheniia-mnozhestvennykh-raka/2026-06-25weekly0.7 https://korshunov.ai/en/article/6465-qehdc-hyperdimensional-computing-based-on-quantum-enhanced-binding-and/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6465-qehdc-vychislenie-giperrazmernosti-na-osnove-kvantovo-usilennogo-sviazyvaniia-i/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6466-asimptoticheskoe-vosstanovlenie-podprostranstva-signala-v-modeliakh-s-softmax/2026-06-25weekly0.7 https://korshunov.ai/en/article/6466-asymptotic-signal-subspace-recovery-in-softmax-attention-models/2026-06-25weekly0.7 https://korshunov.ai/en/article/6467-p4ir-reinforcement-learning-enhances-automated-code-compliance-systems/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6467-p4ir-obuchenie-s-podkrepleniem-povyshaet-effektivnost-sistem-avtomaticheskogo/2026-06-25weekly0.7 https://korshunov.ai/en/article/6464-gara-graph-aware-lora-generation-for-enhancing-llms-on-graph-tasks/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6464-gara-grafovo-osoznannaia-generatsiia-lora-dlia-uluchsheniia-llm-v-grafovykh/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6463-bolshie-iazykovye-modeli-opredeliaiut-prichinnuiu-strukturu-s-pomoshchiu-logiki/2026-06-25weekly0.7 https://korshunov.ai/en/article/6463-llms-determine-causal-structure-via-difference-making-logic/2026-06-25weekly0.7 https://korshunov.ai/en/article/6462-escaping-the-variance-trap-jacobian-free-dynamics-for-root-finding-bilevel/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6462-vykhod-iz-lovushki-dispersii-iakobi-svobodnaia-dinamika-dlia-dvurovnevoi-poiska/2026-06-25weekly0.7 https://korshunov.ai/ru/article/6457-vra-fedsgd-snizhenie-dispersii-v-federativnom-obuchenii-dlia-tiazhelykh-khvostov/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6459-deezer-vnedriaet-sistemu-generatsii-opisanii-pleilistov-na-osnove-bolshikh/2026-06-24weekly0.7 https://korshunov.ai/en/article/6457-vra-fedsgd-variance-reduced-federated-learning-for-heavy-tailed-noise/2026-06-24weekly0.7 https://korshunov.ai/en/article/6459-deezer-deploys-llm-based-music-playlist-captioning-system/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6461-rq-ttsa-raspredelenno-osozdannaia-ustoichivaia-biurovnevaia-optimizatsiia-s/2026-06-24weekly0.7 https://korshunov.ai/en/article/6461-rq-ttsa-distribution-aware-robust-bilevel-optimization-with-quantile-guided/2026-06-24weekly0.7 https://korshunov.ai/en/article/7193-query-on-clustering-nvidia-dgx-spark-and-amd-ryzen-ai-max-395-for-unified-memory/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7193-vopros-o-klasterizatsii-nvidia-dgx-spark-i-amd-ryzen-ai-max-395-dlia-vyvoda-s/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7194-vosstanovlennyi-dok-stantsiia-minisforum-deg1-oculink-dlia-egpu-dostupna-za-59/2026-06-24weekly0.7 https://korshunov.ai/en/article/7195-glm-5-2-on-4x-dgx-spark-reconstructing-missing-build-steps-for-mtp-speculative/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7195-glm-5-2-na-4x-dgx-spark-vosstanovlenie-nedostaiushchikh-shagov-sborki-dlia-mtp/2026-06-24weekly0.7 https://korshunov.ai/en/article/7194-minisforum-deg1-oculink-egpu-dock-refurbished-available-for-59/2026-06-24weekly0.7 https://korshunov.ai/en/article/7190-user-observes-cloud-chatbots-appear-less-intelligent-than-local-models/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7192-colony-obrazovatelnaia-simuliatsiia-mekhanizmov-vnimaniia-llm-s-ispolzovaniem-na/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7190-polzovatel-zamechaet-chto-oblachnye-chat-boty-kazhutsia-menee-intellektualnymi/2026-06-24weekly0.7 https://korshunov.ai/en/article/7192-colony-an-educational-simulation-of-llm-attention-mechanisms-using-agent-based/2026-06-24weekly0.7 https://korshunov.ai/en/article/7185-user-reports-huggingface-charging-for-unused-l40s-compute-in-spaces/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7189-gefen-gotovaia-zamena-adamw-s-zaiavlennym-snizheniem-potrebleniia-pamiati-v-8/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7184-polzovateli-soobshchaiut-o-sboe-dostupa-k-instrumentu-flash-modeli-step-3-7-v/2026-06-24weekly0.7 https://korshunov.ai/en/article/7189-gefen-a-drop-in-replacement-for-adamw-with-claimed-8x-memory-reduction/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7185-polzovatel-soobshchaet-o-vzimanii-platy-za-neispolzuemye-vychisleniia-l40s-v-na/2026-06-24weekly0.7 https://korshunov.ai/en/article/7184-user-reports-step-3-7-flash-model-tool-access-failure-on-huggingchat/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7183-ontologicheskaia-inversiia-perekliuchenie-emotsionalnykh-kontseptsii-llm-s/2026-06-24weekly0.7 https://korshunov.ai/en/article/7181-community-inquiry-on-model-benchmarking-methods/2026-06-24weekly0.7 https://korshunov.ai/en/article/7182-user-inquires-about-organization-rename-process-on-hugging-face/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7181-vopros-soobshchestva-o-metodakh-benchmarkinga-modelei/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7182-polzovatel-sprashivaet-o-protsesse-pereimenovaniia-organizatsii-na-hugging-face/2026-06-24weekly0.7 https://korshunov.ai/en/article/7183-ontological-inversion-flipping-llm-emotional-concepts-via-negative-gain/2026-06-24weekly0.7 https://korshunov.ai/en/article/7191-claude-code-v2-1-191-release-notes/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7188-primechaniia-k-vypusku-llama-index-v0-14-23/2026-06-24weekly0.7 https://korshunov.ai/en/article/7180-qwen3-gemma3-candle-skips-attention-masks-for-equal-length-batches-in-cpu-mode/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7180-qwen3-gemma3-propuskaiut-maski-vnimaniia-dlia-batchei-odinakovoi-dliny-v-rezhime/2026-06-24weekly0.7 https://korshunov.ai/en/article/7188-llama-index-v0-14-23-release-notes/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7191-primechaniia-k-vypusku-claude-code-v2-1-191/2026-06-24weekly0.7 https://korshunov.ai/en/article/7178-v1-14-8a4-release-notes/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7177-reliz-llama-cpp-b9784-optimizatsii-dlia-operatsii-mm-na-osnove-hexagon-i-dlia/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7178-primechaniia-po-vypusku-v1-14-8a4/2026-06-24weekly0.7 https://korshunov.ai/en/article/7177-llama-cpp-release-b9784-hexagon-mm-optimizations-and-cross-platform-binaries/2026-06-24weekly0.7 https://korshunov.ai/en/article/6460-a-differentiable-atari-vcs-for-explainable-ai/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6460-differentsiruemaia-atari-vtss-dlia-obiasnimoi-ii/2026-06-24weekly0.7 https://korshunov.ai/en/article/6458-adar-adaptive-recurrent-message-passing-for-graph-test-time-computing/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6456-modeli-rechi-teksta-latentno-transkribiruiut-rech-v-promezhutochnykh-sloiakh/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6458-adar-adaptivnyi-rekurrentnyi-peredacha-soobshchenii-dlia-grafovykh-vychislenii/2026-06-24weekly0.7 https://korshunov.ai/en/article/6456-speech-text-models-latently-transcribe-speech-in-intermediate-layers/2026-06-24weekly0.7 https://korshunov.ai/en/article/6455-llm-integrated-app-bug-seams-reveal-testing-gaps/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6455-nedostatki-prilozheniia-s-integratsiei-llm-pokazyvaiut-probely-v-testirovanii/2026-06-24weekly0.7 https://korshunov.ai/en/article/7171-sdxl-running-locally-in-browser-on-webgpu-open-source/2026-06-24weekly0.7 https://korshunov.ai/en/article/6452-fed-causaldiff-decoupled-synchronization-for-federated-do-simulation/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6452-fed-causaldiff-razdelennaia-sinkhronizatsiia-dlia-raspredelennoi-simuliatsii/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7171-sdxl-rabotaet-lokalno-v-brauzere-na-webgpu-otkrytyi-iskhodnyi-kod/2026-06-24weekly0.7 https://korshunov.ai/en/article/6453-prompt-side-preprocessing-enhances-edge-ai-accuracy/2026-06-24weekly0.7 https://korshunov.ai/en/article/6454-deep-learning-pipeline-for-sign-language-recognition-and-translation-to-indian/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6453-predvaritelnaia-obrabotka-na-storone-zaprosa-povyshaet-tochnost-edge-ai/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6454-glubokoe-obuchenie-dlia-raspoznavaniia-zhestov-i-perevoda-v-indiiskikh-iazykakh/2026-06-24weekly0.7 https://korshunov.ai/en/article/6449-robust-diffusion-models-via-divergence-induced-weighted-denoising/2026-06-24weekly0.7 https://korshunov.ai/en/article/6450-herta-automated-testing-for-fhe-framework-vulnerabilities/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6450-herta-avtomatizirovannyi-test-na-uiazvimosti-v-freimvorkakh-polnoi-gomomorfnoi/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6451-effekt-nozhnits-rasshirenie-raznoobraziia-ukhudshaet-ustoichivost-perenosa/2026-06-24weekly0.7 https://korshunov.ai/en/article/6451-the-scissors-effect-resize-diversity-hurts-robust-surrogate-transfer/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6449-ustoichivye-diffuzionnye-modeli-cherez-vzveshennoe-udalenie-shuma-osnovannoe-na/2026-06-24weekly0.7 https://korshunov.ai/en/article/6447-introducing-quantum-measurement-temperature-to-stabilize-hybrid-qnn-training/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6448-kadrovyi-ustoichivyi-optimizatsionnyi-freimvork/2026-06-24weekly0.7 https://korshunov.ai/en/article/6448-generative-robust-optimisation-framework/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6447-vvedenie-temperatury-kvantovogo-izmereniia-dlia-stabilizatsii-obucheniia-setei/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6446-glubokaia-setevaia-struktura-dlia-gomogenizatsii-pezoelektricheskikh-kompozitov/2026-06-24weekly0.7 https://korshunov.ai/en/article/6446-deep-material-network-for-homogenization-of-piezoelectric-composites/2026-06-24weekly0.7 https://korshunov.ai/en/article/6445-concept-constrained-prompt-learning-for-few-shot-clip-adaptation/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6445-kontseptualno-ogranichennoe-obuchenie-promptov-dlia-adaptatsii-clip-v-usloviiakh/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6444-statsionarnye-ustoichivye-igry-v-srednem-pole-pri-nesootvetstvii-modelei/2026-06-24weekly0.7 https://korshunov.ai/en/article/6444-stationary-robust-mean-field-games-under-model-mismatches/2026-06-24weekly0.7 https://korshunov.ai/en/article/6441-scalable-bayesian-models-for-stellar-flare-detection/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6443-klassifikatsiia-zadach-bez-obucheniia-dlia-obedineniia-mnogozadachnykh-modelei/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6442-vazhnost-vzveshennaia-na-politicheskaia-distilliatsiia-ustraniaet-smeshchenie/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6441-masshtabiruemye-baiesovskie-modeli-dlia-obnaruzheniia-vspyshek-na-zvezdakh/2026-06-24weekly0.7 https://korshunov.ai/en/article/6443-training-free-task-classification-for-multi-task-model-merging/2026-06-24weekly0.7 https://korshunov.ai/en/article/6442-importance-weighted-on-policy-distillation-addresses-position-bias/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6440-malye-iazykovye-modeli-prevoskhodiat-peredovye-krupnye-iazykovye-modeli-po/2026-06-24weekly0.7 https://korshunov.ai/en/article/7168-i-reverse-engineered-windows-copilot-into-a-free-openai-compatible-api/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7168-ia-razrabotal-windows-copilot-v-vide-besplatnoi-openai-sovmestimoi-api/2026-06-24weekly0.7 https://korshunov.ai/en/article/6440-small-language-models-outperform-frontier-llms-in-relation-extraction/2026-06-24weekly0.7 https://korshunov.ai/en/article/7164-tom-macwright-on-accidental-anonymity-in-job-applications/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7164-tom-makvrait-o-sluchainoi-anonimnosti-v-zaiavkakh-na-rabotu/2026-06-24weekly0.7 https://korshunov.ai/en/article/6439-geometry-aware-online-scheduling-for-llm-serving/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6439-geometriia-orientirovannyi-onlain-raspisyvatel-dlia-obsluzhivaniia-bolshikh/2026-06-24weekly0.7 https://korshunov.ai/en/article/6437-hypothesis-driven-skill-optimization-for-llm-agents/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6437-gipoteza-orientirovannaia-optimizatsiia-navykov-dlia-agentov-na-osnove-modelei/2026-06-24weekly0.7 https://korshunov.ai/en/article/6438-babeljudge-measuring-llm-as-a-judge-reliability-across-languages-and-agent/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6438-babeljudge-otsenka-nadezhnosti-llm-kak-sudi-v-raznykh-iazykakh-i-traektoriiakh/2026-06-24weekly0.7 https://korshunov.ai/en/article/6436-robomme-interference-benchmarking-robot-memory-under-interference/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6436-robomme-interference-otsenka-pamiati-robota-pri-nalichii-pomekh/2026-06-24weekly0.7 https://korshunov.ai/en/article/7163-llama-cpp-releases-b9782-with-new-binaries-and-support/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7162-gemi-3-5-flesh-dobavliaet-funktsiiu-ispolzovaniia-kompiutera/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7163-llama-cpp-vypuskaet-b9782-s-novymi-binarnikami-i-podderzhkoi/2026-06-24weekly0.7 https://korshunov.ai/en/article/7162-gemini-3-5-flash-adds-computer-use-capability/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7158-kto-nibud-eshche-zamechal-chto-vyvody-vllm-khuzhe-chem-v-llama-cpp/2026-06-24weekly0.7 https://korshunov.ai/en/article/6435-flow-annealing-posterior-sampling-for-function-space-regression-and-inverse/2026-06-24weekly0.7 https://korshunov.ai/en/article/7157-sipp-open-source-library-for-in-browser-inference-built-on-llama-cpp/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6435-obratnoe-modelirovanie-posteriornogo-otbora-dlia-regressii-v-prostranstve-i/2026-06-24weekly0.7 https://korshunov.ai/en/article/7158-has-anyone-else-found-vllm-outputs-worse-than-llama-cpp/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7157-sipp-otkrytaia-biblioteka-dlia-inferensa-v-brauzere-postroennaia-na-llama-cpp/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6433-safer-nadezhnaia-adaptatsiia-na-etape-testirovaniia-pri-protivodeistvuiushchikh/2026-06-24weekly0.7 https://korshunov.ai/en/article/6432-sparsity-storage-accuracy-tradeoff-in-parsimoniously-activated-dictionary/2026-06-24weekly0.7 https://korshunov.ai/en/article/6433-safer-reliable-test-time-adaptation-under-adversarial-streams/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6432-transaktsionnyi-kompromiss-mezhdu-redkostiu-stoimostiu-khraneniia-i-tochnostiu-v/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6434-select-to-act-ierarkhicheskaia-rl-s-adaptivnym-rukovodstvom-na-iazyke/2026-06-24weekly0.7 https://korshunov.ai/en/article/6434-select-to-act-hierarchical-rl-with-adaptive-language-guidance/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6431-pervye-tokeny-v-transformerakh-identichnost-iazyka-i-ustoichivost/2026-06-24weekly0.7 https://korshunov.ai/en/article/6431-first-token-broadcasters-in-transformers-language-identity-and-robustness/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6430-otsenka-fizicheskoi-soglasovannosti-v-generatsii-video-bez-ssylok/2026-06-24weekly0.7 https://korshunov.ai/en/article/6428-llm-assisted-label-cleaning-in-chest-ct-dataset/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6428-ochistka-metok-s-ispolzovaniem-bolshikh-iazykovykh-modelei-v-nabore-dannykh-po/2026-06-24weekly0.7 https://korshunov.ai/en/article/6430-reference-free-assessment-of-physical-consistency-in-video-generation/2026-06-24weekly0.7 https://korshunov.ai/en/article/6429-aria-a-causal-aware-framework-for-rescuing-llm-reasoning/2026-06-24weekly0.7 https://korshunov.ai/en/article/6427-hyperadapter-structured-hyperedge-adaptation-for-vision-transformer-fine-tuning/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6429-aria-freimvork-s-uchetom-prichinnosti-dlia-vosstanovleniia-logicheskogo-v-llm/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6427-hyperadapter-strukturirovannaia-adaptatsiia-giperreber-dlia-melkogo-nastroiki/2026-06-24weekly0.7 https://korshunov.ai/en/article/6426-metaps-adaptive-strategy-selection-for-market-agents/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6426-metaps-adaptivnyi-vybor-strategii-dlia-agentov-rynka/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6425-planbench-xl-benchmark-dlia-planirovaniia-ispolzovaniia-instrumentov-na/2026-06-24weekly0.7 https://korshunov.ai/en/article/6424-p4ir-framework-improves-llm-based-code-compliance-accuracy/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6424-kadr-p4ir-povyshaet-tochnost-sobliudeniia-koda-na-osnove-bolshikh-iazykovykh/2026-06-24weekly0.7 https://korshunov.ai/en/article/6425-planbench-xl-benchmark-for-long-horizon-tool-use-planning/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6422-strukturnyi-indeks-bazy-koda-uluchshaet-razreshenie-bez-dopolnitelnoi-stoimosti/2026-06-24weekly0.7 https://korshunov.ai/en/article/6423-gold-points-sniper-self-guided-visual-reasoning-for-fine-grained-action/2026-06-24weekly0.7 https://korshunov.ai/en/article/6422-structural-codebase-index-improves-resolve-without-cost-penalty/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6423-gold-points-sniper-samoreguliruemoe-vizualnoe-myshlenie-dlia-ponimaniia-deistvii/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7150-nvidia-nemo-automodel-uskoriaet-nastroiku-transformerov/2026-06-24weekly0.7 https://korshunov.ai/en/article/7151-introducing-the-ffasr-leaderboard-benchmarking-asr-in-the-real-world/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7151-vvedenie-v-reiting-ffasr-otsenka-asr-v-realnykh-usloviiakh/2026-06-24weekly0.7 https://korshunov.ai/en/article/6421-sciversegym-reinforcement-learning-environment-for-crystal-discovery/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6421-sciversegym-sreda-dlia-obucheniia-s-podkrepleniem-dlia-otkrytiia-kristallov/2026-06-24weekly0.7 https://korshunov.ai/en/article/7150-nvidia-nemo-automodel-speeds-up-transformer-fine-tuning/2026-06-24weekly0.7 https://korshunov.ai/en/article/6420-llms-use-difference-making-logic-to-learn-causal-structure/2026-06-24weekly0.7 https://korshunov.ai/en/article/6419-mmgist-a-comprehensive-multimodal-benchmark-for-2027/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6419-mmgist-kompleksnaia-multimodalnaia-otsenka-dlia-2027-goda/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6420-llms-ispolzuiut-logiku-raznosti-dlia-izucheniia-prichinnoi-struktury/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6418-effektivnye-multimodalnye-modeli-dlia-otsenki-riska-pnevmonii/2026-06-24weekly0.7 https://korshunov.ai/en/article/6418-efficient-multimodal-models-for-pulmonary-embolism-risk-assessment/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6417-dreamuv-konechnyi-potok-dlia-sozdaniia-artistichnykh-razvertyvanii-uv/2026-06-24weekly0.7 https://korshunov.ai/en/article/6417-dreamuv-end-to-end-flow-matching-for-artist-like-uv-unwrapping/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7149-v2-1-190-ispravleniia-oshibok-i-uluchsheniia-nadezhnosti/2026-06-24weekly0.7 https://korshunov.ai/en/article/7149-v2-1-190-bug-fixes-and-reliability-improvements/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6414-raznoobrazie-personazhei-v-istoriiakh-sgenerirovannykh-lingvisticheskimi/2026-06-24weekly0.7 https://korshunov.ai/en/article/6415-self-evolving-cognitive-framework-for-embodied-scientific-intelligence/2026-06-24weekly0.7 https://korshunov.ai/en/article/6414-character-variety-in-llm-generated-stories/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6415-samorazvivaiushchaiasia-kognitivnaia-struktura-dlia-vstroennoi-nauchnoi/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6416-differentsiruemaia-atari-vs-dlia-obiasnimoi-ii/2026-06-24weekly0.7 https://korshunov.ai/en/article/6416-a-differentiable-atari-vcs-for-explainable-ai/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7137-openai-i-broadcom-prezentuiut-chip-dlia-inferensa-na-osnove-bolshikh-iazykovykh/2026-06-24weekly0.7 https://korshunov.ai/en/article/7138-build-a-llm-from-scratch-using-mlx/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7138-sozdanie-llm-s-nulia-s-ispolzovaniem-mlx/2026-06-24weekly0.7 https://korshunov.ai/en/article/7137-openai-and-broadcom-unveil-llm-optimized-inference-chip/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7136-vazhnoe-obnovlenie-dlia-vladeltsev-strix-halo-ot-amd-npu-teper-dostupen-dlia/2026-06-24weekly0.7 https://korshunov.ai/en/article/7136-big-news-for-amd-strix-halo-owners-npu-now-usable/2026-06-24weekly0.7 https://korshunov.ai/en/article/7135-llama-cpp-releases-b9781-with-vulkan-and-multi-platform-support/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7135-llama-cpp-vypuskaet-b9781-s-podderzhkoi-vulkan-i-multiplatformennostiu/2026-06-24weekly0.7 https://korshunov.ai/en/article/6411-llm-integrated-app-bug-seams-reveal-testing-gaps/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6413-prime-otsenka-razresheniia-zaprosov-v-protivorechivykh-instruktsiiakh/2026-06-24weekly0.7 https://korshunov.ai/en/article/6412-factor-enables-adaptive-verification-for-factuality-in-long-form-generation/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6412-factor-obespechivaet-adaptivnuiu-proverku-faktichnosti-v-generatsii-dlinnykh/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6411-nedostatki-prilozheniia-s-integrirovannymi-llm-modeliami-pokazyvaiut-probely-v/2026-06-24weekly0.7 https://korshunov.ai/en/article/6413-prime-evaluating-prompt-resolution-in-conflicting-instructions/2026-06-24weekly0.7 https://korshunov.ai/en/article/6410-vadaorchestra-neurosymbolic-orchestration-of-adaptive-reasoning-workflows/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6410-vadaorchestra-neirosimvolicheskaia-orkestratsiia-adaptivnykh-rabochikh/2026-06-24weekly0.7 https://korshunov.ai/en/article/6409-hi-seg-human-ai-collaboration-for-pulmonary-nodule-segmentation/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6409-hi-seg-kollaboratsiia-cheloveka-i-ii-dlia-segmentatsii-pulmonary-uzlov/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7130-moi-mikro-benchmark-naskolko-khorosho-llms-imitiruiut-povedenie-smachivaniia/2026-06-24weekly0.7 https://korshunov.ai/en/article/7130-my-micro-benchmark-how-good-are-llms-at-simulating-wetting-behaviour/2026-06-24weekly0.7 https://korshunov.ai/en/article/6407-llm-orchestrated-agent-for-soi-directional-coupler-design/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6407-agent-na-osnove-bolshikh-iazykovykh-modelei-dlia-proektirovaniia-kupelei-na-na/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6408-scope-samoadaptivnoe-simvolicheskoe-planirovanie-dlia-otkrytykh-sred/2026-06-24weekly0.7 https://korshunov.ai/en/article/6408-scope-self-adaptive-symbolic-planning-for-open-ended-environments/2026-06-24weekly0.7 https://korshunov.ai/en/article/7128-swiss-federal-supreme-court-evaluates-heretic-for-internal-use/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7128-shveitsarskii-federalnyi-verkhovnyi-sud-otsenivaet-model-heretic-dlia/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7127-novye-elementy-upravleniia-dlia-bezopasnosti-i-dostupa-v-korporativnoi-srede/2026-06-24weekly0.7 https://korshunov.ai/en/article/7127-new-connector-controls-for-enterprise-security-and-access/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7119-prototip-mobilnogo-agenta-aiden/2026-06-24weekly0.7 https://korshunov.ai/en/article/7120-talos-automated-genomic-reanalysis-for-rare-disease-diagnosis/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7120-talos-avtomatizirovannyi-peresmotr-genomnykh-dannykh-dlia-diagnostiki-redkikh/2026-06-24weekly0.7 https://korshunov.ai/en/article/7117-model-hacks-boost-glm5-2-speed-from-2-5-to-over-50-tok-s/2026-06-24weekly0.7 https://korshunov.ai/en/article/7119-aiden-mobile-agent-prototype-in-the-making/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7117-primenenie-khakov-uvelichivaet-skorost-glm5-2-s-25-do-bolee-chem-50-tok-s/2026-06-24weekly0.7 https://korshunov.ai/en/article/6404-prompt-side-preprocessing-enhances-edge-ai-accuracy/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6404-predvaritelnaia-obrabotka-na-storone-zaprosa-povyshaet-tochnost-edge-ai/2026-06-24weekly0.7 https://korshunov.ai/en/article/6405-grounded-scaling-determinism-as-a-core-limit-in-agentic-ai/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6405-zazemlennoe-masshtabirovanie-determinizm-kak-osnovnoe-ogranichenie-v-agentnoi-ii/2026-06-24weekly0.7 https://korshunov.ai/en/article/6406-deep-learning-pipeline-for-sign-language-recognition-and-translation-to-indian/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6406-glubokoe-obuchenie-dlia-raspoznavaniia-zhestov-i-perevoda-na-indiiskie-iazyki/2026-06-24weekly0.7 https://korshunov.ai/en/article/7114-bank-of-korea-report-ai-use-saves-1-hour-weekly-but-no-real-productivity-gain/2026-06-24weekly0.7 https://korshunov.ai/en/article/6402-fed-causaldiff-decoupled-synchronization-for-federated-do-simulation/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6402-fed-causaldiff-raspredelennaia-sinkhronizatsiia-dlia-federativnoi-modelirovaniia/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7114-otchet-narodnogo-banka-korei-ispolzovanie-ii-ekonomit-1-chas-v-nedeliu-no-ne-k/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6403-predstavte-chtoby-obespechit-bezopasnost-v-ierarkhicheskom-obuchenii-s/2026-06-24weekly0.7 https://korshunov.ai/en/article/6403-imagine-to-ensure-safety-in-hierarchical-reinforcement-learning/2026-06-24weekly0.7 https://korshunov.ai/en/article/6400-generative-robust-optimisation-framework/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6399-gazer-obnaruzhenie-oshibok-semantiki-bez-obucheniia-dlia-avtokorrektiruiushchikh/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6401-degradatsiia-upravleniia-v-agentakh-na-dlinnykh-gorizontakh-llm/2026-06-24weekly0.7 https://korshunov.ai/en/article/6401-governance-decay-in-long-horizon-llm-agents/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6400-kadrovaia-ustoichivaia-optimizatsiia/2026-06-24weekly0.7 https://korshunov.ai/en/article/6399-gazer-training-free-semantic-correction-for-autoregressive-visual-models/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7113-openai-i-broadcom-prezentuiut-chip-dlia-inferensa-bolshikh-iazykovykh-modelei/2026-06-24weekly0.7 https://korshunov.ai/en/article/7113-openai-and-broadcom-unveil-llm-optimized-inference-chip/2026-06-24weekly0.7 https://korshunov.ai/en/article/7110-nex-n2-mini-ultra-uncensored-heretic-model-released/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6398-macagentbench-zapuskaet-benchmark-iskusstvennogo-intellekta-dlia-agentov-macos/2026-06-24weekly0.7 https://korshunov.ai/en/article/6398-macagentbench-launches-macos-ai-agent-benchmark/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7110-vypushchen-model-nex-n2-mini-ultra-uncensored-heretic/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7102-veb-interfeis-llama-cpp-dobavliaet-optsionalnoe-vypolnenie-javascript-cherez-web/2026-06-24weekly0.7 https://korshunov.ai/en/article/7102-llama-cpp-web-ui-adds-optional-javascript-execution-via-web-workers/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7103-qwen-agentworld-35b-a3b-dlia-programmirovaniia/2026-06-24weekly0.7 https://korshunov.ai/en/article/7103-qwen-agentworld-35b-a3b-for-coding/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7104-proverka-nadezhnosti-dvukh-gpu-iavliaetsia-li-eto-vygodnoi-pokupkoi/2026-06-24weekly0.7 https://korshunov.ai/en/article/7104-dual-gpu-sanity-check-is-this-a-smart-buy/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6397-multimodalnyi-tsepochka-myslei-vozmozhnosti-i-ogranicheniia/2026-06-24weekly0.7 https://korshunov.ai/en/article/6397-multimodal-chain-of-thought-capabilities-and-limitations/2026-06-24weekly0.7 https://korshunov.ai/en/article/6395-smartsdg-pipeline-enhances-syn-to-real-object-detection/2026-06-24weekly0.7 https://korshunov.ai/en/article/6396-concept-constrained-prompt-learning-for-few-shot-clip-adaptation/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6395-smartsdg-pipeline-enhances-syn-to-real-object-detection/2026-06-24weekly0.7 https://korshunov.ai/en/article/6394-context-aware-distillation-and-ablation-for-text2dsl/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6394-kontekstualno-osoznannoe-distillirovanie-i-ablatsiia-dlia-text2dsl/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6396-kontseptualno-ogranichennoe-obuchenie-promptov-dlia-maloshotnogo-adaptirovaniia/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7098-kakie-instrumenty-ispolzuiut-liudi-dlia-otsenki-vram-i-ram-dlia-lokalnykh-llm/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7099-gemma-4-26ba4b-stranno-polezen-pri-iq3-s/2026-06-24weekly0.7 https://korshunov.ai/en/article/7099-gemma-4-26ba4b-surprisingly-usable-at-iq3-s/2026-06-24weekly0.7 https://korshunov.ai/en/article/7098-what-tools-do-people-use-to-estimate-vram-and-ram-for-local-llms/2026-06-24weekly0.7 https://korshunov.ai/en/article/6393-cwe-level-generalisation-in-syscall-based-hids/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6393-universalizatsiia-na-urovne-cwe-v-sistemakh-nabliudeniia-za-sistemnymi-vyzovami/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6391-vesovoe-nastroika-na-poli-distilliatsii-reshaet-smeshchenie-pozitsii/2026-06-24weekly0.7 https://korshunov.ai/en/article/6392-text2dsl-llm-based-code-generation-for-domain-specific-languages/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6392-text2dsl-generatsiia-koda-na-iazykakh-s-opredelennoi-oblastiu-primeneniia-na/2026-06-24weekly0.7 https://korshunov.ai/en/article/7095-baidu-s-unlimited-ocr-transcribes-dozens-of-pages-in-one-forward-pass/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7095-model-unlimited-ocr-ot-baidu-perevodit-desiatki-stranits-za-odin-prokhod/2026-06-24weekly0.7 https://korshunov.ai/en/article/6391-importance-weighted-on-policy-distillation-addresses-position-bias/2026-06-24weekly0.7 https://korshunov.ai/en/article/6335-optimal-subsampling-in-rkhs-for-supervised-learning/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6390-paperclaw-avtonomnye-issledovaniia-s-vovlecheniem-cheloveka-na-etape-utochneniia/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6335-optimalnaia-subsempling-v-rkhs-dlia-zadachi-nadzora/2026-06-24weekly0.7 https://korshunov.ai/en/article/6334-reward-free-pretraining-for-reinforcement-learning-via-occupancy-coverage/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6334-nagrada-bezopasnoe-predobuchenie-dlia-obucheniia-s-voznagrazhdeniem-cherez/2026-06-24weekly0.7 https://korshunov.ai/en/article/6390-paperclaw-autonomous-research-with-human-in-the-loop-refinement/2026-06-24weekly0.7 https://korshunov.ai/en/article/6332-topological-neural-dynamics-neuron-wise-sequence-modeling/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6332-topologicheskie-neironnye-dinamiki-modelirovanie-posledovatelnostei-po-neironam/2026-06-24weekly0.7 https://korshunov.ai/en/article/6330-taser-task-differentiated-skill-expansion-for-heterogeneous-continual-learning/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6330-taser-rasshirenie-navykov-na-osnove-zadach-dlia-nepreryvnogo-obucheniia-v/2026-06-24weekly0.7 https://korshunov.ai/en/article/6333-teanet-improves-few-shot-learning-in-vibrational-spectroscopy/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6333-teanet-uluchshaet-obuchenie-s-nebolshim-kolichestvom-primerov-v-vibratsionnoi/2026-06-24weekly0.7 https://korshunov.ai/en/article/6331-nasdaq-normalized-observation-space-dynamics-augmented-q-learning/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6331-nasdaq-usrednennaia-dinamika-nabliudenii-s-usileniem-q-obucheniia/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7094-qwen3-6-27b-v-vllm-dumber-chem-v-llama-cpp/2026-06-24weekly0.7 https://korshunov.ai/en/article/7094-qwen3-6-27b-more-dumb-in-vllm-compared-to-llama-cpp/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6326-medts-ttt-obuchenie-v-protsesse-testirovaniia-dlia-meditsinskikh-vremennykh/2026-06-24weekly0.7 https://korshunov.ai/en/article/6329-diagnostics-for-morl-policy-selection/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6329-diagnostika-dlia-vybora-politiki-v-morl/2026-06-24weekly0.7 https://korshunov.ai/en/article/6326-medts-ttt-test-time-training-for-medical-time-series/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7090-kalm-reranker-v1-bystryi-i-effektivnyi-pereuporiadochivatel-dokumentov/2026-06-24weekly0.7 https://korshunov.ai/en/article/7090-kalm-reranker-v1-fast-and-efficient-document-reranking/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6328-obnaruzhenie-anomalii-bez-nadzora-s-pomoshchiu-rezervuarnykh-kompiuterov/2026-06-24weekly0.7 https://korshunov.ai/en/article/6327-sea-scan-ml-based-dark-vessel-detection-with-weak-supervision/2026-06-24weekly0.7 https://korshunov.ai/en/article/6328-unsupervised-anomaly-detection-with-reservoir-computers/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6327-sea-scan-obnaruzhenie-temnykh-sudov-na-osnove-mashinnogo-obucheniia-s-slabogo/2026-06-24weekly0.7 https://korshunov.ai/en/article/7087-llama-cpp-updates-granite-speech-lfm2-5-colbert-models-vulkan-backend/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7087-obnovleniia-llama-cpp-modeli-granite-speech-lfm2-5-colbert-uluchsheniia-v-vulkan/2026-06-24weekly0.7 https://korshunov.ai/en/article/7089-vulkan-shaders-gen-now-fails-build-on-shader-compilation-errors/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7089-vulkan-shaders-gen-teper-preryvaet-sborku-pri-oshibkakh-kompiliatsii-sheiderov/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6325-remont-grafa-ramanuia-snizhaet-perenasyshchenie-v-gnn/2026-06-24weekly0.7 https://korshunov.ai/en/article/6325-ramanujan-graph-rewiring-alleviates-gnn-over-squashing/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6324-rezervuarnoe-vychislenie-dlia-obrabotki-zvukovykh-signalov-bez-izvlecheniia/2026-06-24weekly0.7 https://korshunov.ai/en/article/6324-reservoir-computing-for-feature-free-audio-signal-processing/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6322-modeli-transformer-chrezvychaino-chuvstvitelny-k-shumu-v-dannykh-o-traektorii/2026-06-24weekly0.7 https://korshunov.ai/en/article/6322-transformer-models-highly-sensitive-to-noisy-data-in-trajectory-prediction/2026-06-24weekly0.7 https://korshunov.ai/en/article/6321-reward-petri-net-interpretation-of-temporal-behavior-trees/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6321-interpretatsiia-vremennykh-derevev-povedeniia-s-pomoshchiu-reward-petri-net/2026-06-24weekly0.7 https://korshunov.ai/en/article/6323-dataclaw0-agentic-tailoring-of-multimodal-data-from-raw-streams/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6323-dataclaw0-agentnaia-nastroika-multimodalnykh-dannykh-iz-iskhodnykh-potokov/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6320-kadrovaia-sistema-na-osnove-otkrytykh-dannykh-opredeliaet-topologiiu-gorodskoi/2026-06-24weekly0.7 https://korshunov.ai/en/article/6320-open-data-framework-identifies-urban-power-grid-topology/2026-06-24weekly0.7 https://korshunov.ai/en/article/6317-graph-of-differences-for-anatomy-structured-medreid/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6318-upravlenie-prognozirovaniem-remonta-s-ispolzovaniem-mnogourovnevogo-vnimaniia-i/2026-06-24weekly0.7 https://korshunov.ai/en/article/6319-sohet-transformer-for-heterogeneous-event-streams/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6319-sohet-transformator-dlia-geterogennykh-potokov-sobytii/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6317-graf-raznostei-dlia-anatomicheski-strukturirovannoi-meditsinskoi-identifikatsii/2026-06-24weekly0.7 https://korshunov.ai/en/article/6318-predictive-repair-management-using-multi-head-attention-and-online-learning/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6316-neironnyi-kodek-deistvii-dlia-modelei-vizualno-iazykovykh-deistvii/2026-06-24weekly0.7 https://korshunov.ai/en/article/6314-functional-orthogonality-ensures-identifiability-in-unsupervised-disentanglement/2026-06-24weekly0.7 https://korshunov.ai/en/article/6316-neural-action-codec-for-vision-language-action-models/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6314-funktsionalnaia-ortogonalnost-obespechivaet-identifitsiruemost-v-neperemennoi/2026-06-24weekly0.7 https://korshunov.ai/en/article/6315-triz-inspired-text-to-cad-framework-enhances-creative-design/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6315-kadrovaia-sistema-na-osnove-triz-uluchshaet-kreativnyi-dizain/2026-06-24weekly0.7 https://korshunov.ai/en/article/6311-cat-translate-compact-japanese-english-translation-models/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6312-atomarnye-iazykovye-modeli-ponimaiut-i-generiruiut-materialy/2026-06-24weekly0.7 https://korshunov.ai/en/article/6312-atomistic-language-models-understand-and-generate-materials/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6313-vla-fail-legkaia-sistema-obnaruzheniia-sboev-dlia-modelei-vizualno-iazykovykh/2026-06-24weekly0.7 https://korshunov.ai/en/article/6313-vla-fail-lightweight-failure-detection-for-vision-language-action-models/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6311-cat-translate-kompaktnye-modeli-perevoda-iaponskogo-angliiskogo-iazyka/2026-06-24weekly0.7 https://korshunov.ai/en/article/6310-adualvuot-heterogeneous-latent-space-alignment-for-unsupervised-domain/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6309-freimvork-ldt-frl-dlia-kiber-ustoichivykh-iomt/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6310-adualvuot-geterogennaia-algebraicheskaia-sinkhronizatsiia-prostranstva-skrytykh/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6308-dvoinoe-vnimanie-konvoliutsionnykh-ekspertov-dlia-zapolneniia-redkikh-tenzorov/2026-06-24weekly0.7 https://korshunov.ai/en/article/6308-dual-attention-convolution-experts-for-sparse-tensor-completion/2026-06-24weekly0.7 https://korshunov.ai/en/article/6309-ldt-frl-framework-for-cyber-resilient-iomt/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6306-obshchie-kodery-dlia-modulnogo-reliatsionnogo-glubokogo-obucheniia/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7084-reliz-llama-cpp-b9777-dobavliaet-novye-modeli-i-binarniki-dlia-raznykh-platform/2026-06-24weekly0.7 https://korshunov.ai/en/article/6307-central-limit-theorem-for-averaged-adam-optimizer/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6307-tsentralnaia-predelnaia-teorema-dlia-srednego-optimizatora-adam/2026-06-24weekly0.7 https://korshunov.ai/en/article/7084-llama-cpp-release-b9777-adds-new-models-and-cross-platform-binaries/2026-06-24weekly0.7 https://korshunov.ai/en/article/6306-universal-encoders-for-modular-relational-deep-learning/2026-06-24weekly0.7 https://korshunov.ai/en/article/6304-fast-turboquant-multiplier-free-vector-quantization/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6305-kadr-bipc-uskoriaet-optimizatsiiu-smeshannykh-tselykh-chisel-s-pomoshchiu/2026-06-24weekly0.7 https://korshunov.ai/en/article/6305-bipc-framework-accelerates-mixed-integer-optimization-with-machine-learning/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6304-fast-turboquant-multiplikatornaia-bezopasnaia-vektornaia-kvantovanie/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6301-ascii-art-pozvoliaet-tekstovym-llm-kontrolirovat-sistemy-vla/2026-06-24weekly0.7 https://korshunov.ai/en/article/6302-native-space-pipelines-outperform-template-space-in-subcortical-segmentation/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6302-pipelines-na-osnove-prostranstva-subekta-prevoskhodiat-analogi-v-prostranstve/2026-06-24weekly0.7 https://korshunov.ai/en/article/6301-ascii-art-enables-text-only-llms-to-control-vla-systems/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6303-postobucheniia-uluchsheniia-rechi-s-pertseptualnymi-voznagrazhdeniiami/2026-06-24weekly0.7 https://korshunov.ai/en/article/6303-post-training-speech-enhancement-with-perceptual-rewards/2026-06-24weekly0.7 https://korshunov.ai/en/article/6300-deep-learning-fuses-satellite-data-with-meteorological-features-for-soil/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6299-nesootvetstvie-obucheniia-s-pomoshchiu-protivorechivykh-dannykh-dlia-nelineinykh/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7082-novyi-evropeiskii-model-ai-domyn-budet-imet-400-milliardov-parametrov/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6300-glubokoe-obuchenie-obediniaet-dannye-sputnikov-s-meteorologicheskimi-dlia-pochvy/2026-06-24weekly0.7 https://korshunov.ai/en/article/7082-new-eu-ai-model-domyn-will-be-400b-parameters/2026-06-24weekly0.7 https://korshunov.ai/en/article/6299-adversarial-training-equivalence-fails-for-nonlinear-models/2026-06-24weekly0.7 https://korshunov.ai/en/article/6295-js-divergence-enhances-grpo-autoregressive-text-to-image-alignment/2026-06-24weekly0.7 https://korshunov.ai/en/article/6298-machine-learning-model-predicts-high-risk-colorectal-polyps-in-african-americans/2026-06-24weekly0.7 https://korshunov.ai/en/article/7081-v0-24-0rc1-fix-topk-histogram-build-on-sm75/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6298-mashinnoe-obuchenie-predskazyvaet-vysokoriskovye-polipy-kishechnika-u/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7081-v0-24-0rc1-ispravlenie-sborki-gistogrammy-topk-na-sm75/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6295-js-razdelenie-povyshaet-avtokorreliatsionnuiu-sinkhronizatsiiu-teksta-i-v-grpo/2026-06-24weekly0.7 https://korshunov.ai/en/article/6297-decoupling-declarative-and-procedural-knowledge-in-vision-language-action-models/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6296-glubokoe-obuchenie-s-parallelnoi-vremennoi-slozhnostiu-o-log-n/2026-06-24weekly0.7 https://korshunov.ai/en/article/6296-deep-learning-with-o-log-n-parallel-time-complexity/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6297-razdelenie-deklarativnykh-i-protsedurnykh-znanii-v-modeliakh-vizualno-iazykovykh/2026-06-24weekly0.7 https://korshunov.ai/en/article/6292-analytic-policy-gradients-for-efficient-continuous-control/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6292-analiticheskie-gradienty-politik-dlia-effektivnogo-nepreryvnogo-upravleniia/2026-06-24weekly0.7 https://korshunov.ai/en/article/6294-privacy-preserving-federated-temporal-graph-learning-for-cyber-resilient-iomt/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6294-anonimizirovannoe-raspredelennoe-obuchenie-vremennykh-grafov-dlia-kiber-sistem/2026-06-24weekly0.7 https://korshunov.ai/en/article/7078-best-model-for-local-usage-and-working-on-unity-with-mcp-at-12-gb-vram/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7078-luchshaia-model-dlia-lokalnoi-raboty-i-raboty-v-unity-s-mcp-pri-12-gb-vram/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7077-wav2vec2-i-wavlm-klassifikatory-audio-zastriali-na-33-tochnosti/2026-06-24weekly0.7 https://korshunov.ai/en/article/7077-wav2vec2-and-wavlm-audio-classifier-stuck-at-33-accuracy/2026-06-24weekly0.7 https://korshunov.ai/en/article/7079-i-built-an-mcp-server-in-go-for-ai-agents-200-lines-tutorial/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7079-ia-sozdal-server-mcp-na-go-dlia-ii-agentov-200-strokovyi-tiutorial/2026-06-24weekly0.7 https://korshunov.ai/en/article/6291-goma-achieves-first-stochastic-convergence-guarantee-for-variational/2026-06-24weekly0.7 https://korshunov.ai/en/article/6293-muon-optimizer-power-limits-and-a-river-valley-theory/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6291-goma-dostigaet-pervogo-garantii-stokhasticheskoi-skhodimosti-dlia-variatsionnykh/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6293-optimizator-miuon-sila-predely-i-teoriia-rechnoi-doliny/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7073-kto-oplachivaet-razrabotku-lokalnykh-modelei-iazykovoi-obrabotki/2026-06-24weekly0.7 https://korshunov.ai/en/article/7073-who-s-paying-for-local-llm-work/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7069-llama-server-avariino-ostanavlivaetsia-pri-ispolzovanii-prompta-vstavlennogo-kak/2026-06-24weekly0.7 https://korshunov.ai/en/article/7069-llama-server-crashes-with-pasted-as-file-prompt-for-image-data-extraction/2026-06-24weekly0.7 https://korshunov.ai/en/article/7071-pcie-5-0-16x-split-into-2x8-with-riser-cable/2026-06-24weekly0.7 https://korshunov.ai/en/article/7070-qwen-releases-qwen-agentworld-397b-a17b-model/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7070-qwen-predstavil-model-qwen-agentworld-397b-a17b/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7071-pcie-5-0-16x-razdeliaetsia-na-2x8-s-pomoshchiu-kabelia-razema/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7068-unlimited-ocr-teper-dostupen-na-modelscope/2026-06-24weekly0.7 https://korshunov.ai/en/article/7066-qwen-releases-35b-parameter-moe-for-agent-environment-simulation/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7066-qwen-vypustil-moe-na-35-milliardov-parametrov-dlia-imitatsii-sredy-agenta/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7067-repozitorii-github-qwen-agentworld-dlia-iazykovykh-mirovykh-modelei/2026-06-24weekly0.7 https://korshunov.ai/en/article/7067-github-repository-qwen-agentworld-for-language-world-models/2026-06-24weekly0.7 https://korshunov.ai/en/article/7068-unlimited-ocr-is-now-on-modelscope/2026-06-24weekly0.7 https://korshunov.ai/en/article/7062-chinese-supercomputer-displaces-us-machines-as-world-s-fastest-for-first-time/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7062-kitaiskii-superkompiuter-vytesniaet-amerikanskie-mashiny-i-stanovitsia-samym-v-s/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7061-zakonoproekt-o-prinuditelnom-vnedrenii-otslezhivaniia-mestopolozheniia-chipov-ii/2026-06-24weekly0.7 https://korshunov.ai/en/article/7061-bill-to-mandate-ai-chip-location-tracking-gains-industry-support/2026-06-24weekly0.7 https://korshunov.ai/en/article/7010-studying-user-mental-models-in-speech-translation/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7010-issledovanie-psikhicheskikh-modelei-polzovatelei-v-perevode-rechi/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7009-parapairaudiobench-benchmark-dlia-otsenki-paralingvisticheskikh-kharakteristik/2026-06-24weekly0.7 https://korshunov.ai/en/article/7009-parapairaudiobench-benchmark-for-paralinguistic-speech-evaluation/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7007-dream-avtoregressivnoe-obuchenie-dlia-plotnykh-vektorov-poiska/2026-06-24weekly0.7 https://korshunov.ai/en/article/7008-ai-pave-br-llm-based-pave-for-brazilian-e-commerce/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7008-ai-pave-br-llm-base-pave-dlia-brazilskogo-elektronnogo-kommertsii/2026-06-24weekly0.7 https://korshunov.ai/en/article/7007-dream-autoregressive-training-for-dense-retrieval-embeddings/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7006-vypushchena-versiia-v0-1-cn-newstts-bench/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7004-candle-legkaia-deduplikatsiia-shumov-na-arabskom-iazyke-s-ispolzovaniem-ctc/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7005-razdelenie-zadachi-dlia-effektivnoi-annotatsii/2026-06-24weekly0.7 https://korshunov.ai/en/article/7005-task-decomposition-for-efficient-annotation/2026-06-24weekly0.7 https://korshunov.ai/en/article/7004-candle-lightweight-arabic-noise-deduplication-via-ctc/2026-06-24weekly0.7 https://korshunov.ai/en/article/7006-cn-newstts-bench-v0-1-released/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7003-posteriornaia-korrektsiia-bystraia-generatsiia-iazyka-cherez-liubye-poriadki/2026-06-24weekly0.7 https://korshunov.ai/en/article/7001-micro-transaction-markets-for-verified-product-info-in-agentic-e-commerce/2026-06-24weekly0.7 https://korshunov.ai/en/article/7003-posterior-refinement-fast-language-generation-via-any-order-flow-maps/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7002-gotovy-li-my-k-agentno-orientirovannoi-sisteme-pamiati/2026-06-24weekly0.7 https://korshunov.ai/en/article/7002-are-we-ready-for-an-agent-native-memory-system/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7001-rynki-mikro-tranzaktsii-dlia-proverennoi-informatsii-o-produktakh-v-agentnoi/2026-06-24weekly0.7 https://korshunov.ai/en/article/6999-l3cube-mahapos-marathi-pos-tagging-dataset-and-bert-models/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6999-l3cube-mahapos-dataset-maratkhi-i-modeli-bert/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6998-vybor-obuchaiushchikh-dannykh-s-uchetom-kachestva-dlia-nauchnogo-sinteza/2026-06-24weekly0.7 https://korshunov.ai/en/article/6998-quality-aware-training-data-selection-for-scientific-summarization/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7000-sherloc-strukturirovannaia-lokalizatsiia-diagnostiki-dlia-agentov-koda/2026-06-24weekly0.7 https://korshunov.ai/en/article/7000-sherloc-structured-diagnostic-localization-for-code-repair-agents/2026-06-24weekly0.7 https://korshunov.ai/en/article/6997-match-task-to-objective-framework-for-encoder-decoder-models/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6997-sootvetstvie-zadachi-i-ramochnoi-struktury-dlia-modelei-kodirovshchikov/2026-06-24weekly0.7 https://korshunov.ai/en/article/6991-linguistic-fingerprints-reveal-tang-poets-regional-origins/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6991-lingvisticheskie-otpechatki-raskryvaiut-regionalnye-proiskhozhdeniia-tanskikh/2026-06-24weekly0.7 https://korshunov.ai/en/article/6990-first-large-scale-analysis-of-algorithm-co-occurrence-networks/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6990-pervyi-masshtabnyi-analiz-setei-sovmestnogo-poiavleniia-algoritmov/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7063-reliz-openai-agents-python-v0-17-7/2026-06-24weekly0.7 https://korshunov.ai/en/article/6988-lora-monitor-calibration-fails-with-top-1-in-diffusion-lms/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6988-oshibki-kalibrovki-lora-monitor-pri-ispolzovanii-top-1-v-diffuzionnykh-modeliakh/2026-06-24weekly0.7 https://korshunov.ai/en/article/6989-porter-language-grounded-event-representations-for-portable-ehr-foundation/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6989-porter-iazykovo-obosnovannye-predstavleniia-sobytii-dlia-portativnykh-modelei/2026-06-24weekly0.7 https://korshunov.ai/en/article/7063-release-of-openai-agents-python-v0-17-7/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6985-metis-sviazyvanie-pamiati-teksta-i-koda-dlia-samoevoliutsionnykh-agentov/2026-06-24weekly0.7 https://korshunov.ai/en/article/6985-metis-bridging-text-and-code-memory-for-self-evolving-agents/2026-06-24weekly0.7 https://korshunov.ai/en/article/6986-interaligner-progressive-alignment-for-asr/2026-06-24weekly0.7 https://korshunov.ai/en/article/6987-holistic-data-scheduler-for-llm-pre-training-via-multi-objective-reinforcement/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6987-tselostnyi-planirovshchik-dannykh-dlia-predobucheniia-iazykovykh-modelei-s-s/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6986-interaligner-progressivnaia-sinkhronizatsiia-dlia-audio-raspoznavaniia/2026-06-24weekly0.7 https://korshunov.ai/en/article/6984-medbench-v5-dynamic-benchmark-for-clinical-ai/2026-06-24weekly0.7 https://korshunov.ai/en/article/6982-core-brew-llr-based-soft-decoding-for-robust-multi-bit-llm-watermarking/2026-06-24weekly0.7 https://korshunov.ai/en/article/6983-behaviorbench-launches-benchmark-for-behavioral-ai-models/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6982-core-brew-miagkii-dekoding-na-osnove-log-veroiatnostnykh-otnoshenii-dlia-vstavki/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6983-behaviorbench-zapuskaet-benchmark-dlia-povedencheskikh-ii-modelei/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6984-medbench-v5-dinamicheskii-benchmark-dlia-klinicheskoi-ii/2026-06-24weekly0.7 https://korshunov.ai/en/article/6979-agon-autonomous-research-system-via-prompt-economy/2026-06-24weekly0.7 https://korshunov.ai/en/article/6980-digi-turbine-a-reliability-aware-pinn-benchmark-for-offshore-wind-monitoring/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6980-digi-turbine-sinteticheskii-benchmark-s-uchetom-nadezhnosti-dlia-monitoringa/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6979-agon-avtonomnaia-sistema-issledovanii-cherez-ekonomiku-zaprosov/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6981-fond-panini-dlia-obrabotki-indiiskikh-iazykov/2026-06-24weekly0.7 https://korshunov.ai/en/article/6981-paninian-foundation-for-indic-language-processing/2026-06-24weekly0.7 https://korshunov.ai/en/article/6976-dialogue-to-discovery-attribute-aware-preference-elicitation/2026-06-24weekly0.7 https://korshunov.ai/en/article/6977-recare-robust-erasure-for-co-occurring-retained-concepts-in-diffusion-unlearning/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6976-dialog-k-otkrytiiu-elicitatsiia-predpochtenii-s-uchetom-atributov/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6978-evoliutsiia-aspektnoi-emotsionalnoi-otsenki-v-mnogoetapnykh-vzaimnykh-otsenkakh/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6977-recare-ustoichivoe-udalenie-dlia-soputstvuiushchikh-sokhraniaiushchikhsia-v/2026-06-24weekly0.7 https://korshunov.ai/en/article/6978-aspect-based-sentiment-evolution-in-multi-round-peer-reviews/2026-06-24weekly0.7 https://korshunov.ai/en/article/6971-caliber-calibrating-confidence-before-and-after-reasoning-in-language-models/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6974-dekogerentsiia-kak-zashchita-v-kvantovykh-neironnykh-setiakh-dlia-obnaruzheniia/2026-06-24weekly0.7 https://korshunov.ai/en/article/6974-decoherence-as-defence-in-quantum-neural-networks-for-intrusion-detection/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6975-mmed-bench-ir-multimaliarnyi-meditsinskii-benchmark-po-poisku/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6971-caliber-kalibrovka-uverennosti-do-i-posle-rassuzhdeniia-v-modeliakh-iazyka/2026-06-24weekly0.7 https://korshunov.ai/en/article/6975-mmed-bench-ir-a-multilingual-medical-retrieval-benchmark/2026-06-24weekly0.7 https://korshunov.ai/en/article/6973-surgellm-task-aware-feature-gating-with-class-balanced-normalization/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6973-surgellm-zadacha-orientirovannaia-geitingovaia-ficha-s-klassovo-sbalansirovannoi/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6972-plokhie-zaprosy-privodiat-k-kollapsu-modeli-i-oshibkam/2026-06-24weekly0.7 https://korshunov.ai/en/article/6972-bad-prompts-cause-model-collapse-and-mistakes/2026-06-24weekly0.7 https://korshunov.ai/en/article/6969-prague-dependency-treebank-consolidated-2-0-released/2026-06-24weekly0.7 https://korshunov.ai/en/article/7013-llama-cpp-release-b9776-adds-vulkan-and-multiple-hardware-support/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6969-pragskii-derevoobraznyi-korpus-obnovlennyi-do-versii-2-0/2026-06-24weekly0.7 https://korshunov.ai/en/article/6970-avoc-retrieval-inspired-token-compression-for-long-form-audio-video/2026-06-24weekly0.7 https://korshunov.ai/ru/article/7013-llama-cpp-release-b9776-dobavliaet-podderzhku-vulkan-i-neskolkikh-apparatnykh/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6970-avoc-ispolzovanie-metoda-izvlecheniia-dlia-szhatiia-tokenov-v-ponimanii-dlinnykh/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6966-petra-dataset-i-paiplain-dlia-adaptatsii-tekstov-v-neftegazovoi-inzhenerii/2026-06-24weekly0.7 https://korshunov.ai/en/article/6967-ud-czech-pdtc-a-large-and-genre-rich-treebank-in-universal-dependencies/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6968-modeli-transformerov-arkhitektury-primeneniia-i-kriticheskii-analiz/2026-06-24weekly0.7 https://korshunov.ai/en/article/6968-transformer-models-architectures-applications-and-critical-assessment/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6967-ud-czech-pdtc-bolshoi-i-zhanrovo-raznoobraznyi-korpus-v-universalnykh/2026-06-24weekly0.7 https://korshunov.ai/en/article/6966-petra-dataset-and-pipeline-for-petroleum-engineering-text-adaptation/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6965-pozitsionnaia-markirovka-smysla-arabsko-angliiskogo-slovaria-s-pomoshchiu/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6963-computefhe-biblioteka-obshchego-naznacheniia-dlia-vychislenii-s-obespecheniem/2026-06-24weekly0.7 https://korshunov.ai/en/article/6963-computefhe-a-privacy-preserving-general-purpose-computation-library/2026-06-24weekly0.7 https://korshunov.ai/en/article/6964-morfflex-managing-rich-morphology-in-czech/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6964-morfflex-upravlenie-bogatoi-morfologiei-na-cheshskom-iazyke/2026-06-24weekly0.7 https://korshunov.ai/en/article/6965-pos-tagging-of-arabic-english-dictionary-senses-via-wordnet/2026-06-24weekly0.7 https://korshunov.ai/en/article/6962-stability-of-prompt-ranking-in-llm-evaluation/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6961-autospecner-dataset-dlia-detalizirovannogo-raspoznavaniia-nazvanii-v-avtomobilei/2026-06-24weekly0.7 https://korshunov.ai/en/article/6961-autospecner-fine-grained-ner-dataset-for-vehicle-specifications/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6962-stabilnost-ranzhirovaniia-promptov-v-otsenke-llm/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6960-era-llm-benchmark-dlia-logicheskogo-myshleniia-i-diplomatii-v-llm/2026-06-24weekly0.7 https://korshunov.ai/en/article/6960-age-of-llm-benchmark-for-llm-reasoning-and-diplomacy/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6958-kadr-edv-obespechivaet-nadezhnoe-obuchenie-opyta-dlia-agentnykh-sistem/2026-06-24weekly0.7 https://korshunov.ai/en/article/6959-extractconf-multi-signal-confidence-engine-for-llm-document-extraction/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6957-baiesovskii-kontrol-dlia-agentov-kodirovaniia/2026-06-24weekly0.7 https://korshunov.ai/en/article/6957-bayesian-control-for-coding-agents/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6959-extractconf-mnogosignalnyi-dvigatel-uverennosti-dlia-izvlecheniia-dokumentov-iz/2026-06-24weekly0.7 https://korshunov.ai/en/article/6958-edv-framework-enables-reliable-experience-learning-for-agentic-systems/2026-06-24weekly0.7 https://korshunov.ai/en/article/6955-african-language-tokenization-penalty-in-frontier-llms/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6956-llm-osnovannaia-dvukhetapnaia-transformatornaia-model-dlia-diagnostiki/2026-06-24weekly0.7 https://korshunov.ai/en/article/6954-uol-idem-submits-l1-aware-vocabulary-prediction-model/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6955-shtraf-na-tokenizatsiiu-afrikanskikh-iazykov-v-peredovykh-modeliakh-iazykov/2026-06-24weekly0.7 https://korshunov.ai/en/article/6956-llm-based-two-stage-transformer-for-bearing-fault-diagnosis/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6954-uol-idem-predstavliaet-model-predskazaniia-slovarnogo-urovnia-s-uchetom-l1/2026-06-24weekly0.7 https://korshunov.ai/en/article/6951-agora-benchmark-for-agentic-workplace-document-reasoning/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6953-radar-ii-model-uluchshaet-diagnostiku-redkikh-zabolevanii/2026-06-24weekly0.7 https://korshunov.ai/en/article/6953-radar-ai-model-improves-rare-disease-diagnosis/2026-06-24weekly0.7 https://korshunov.ai/en/article/6952-poster-exploring-audio-based-scam-detection-in-turkish/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6951-agora-benchmark-dlia-agentnogo-analiza-dokumentov-v-rabochei-srede/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6952-poster-issledovanie-obnaruzheniia-moshennicheskikh-zvonkov-na-osnove-audio-v/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6949-mezhiazykovoe-issledovanie-parametricheskikh-znanii/2026-06-24weekly0.7 https://korshunov.ai/en/article/6949-cross-lingual-exploration-for-parametric-knowledge/2026-06-24weekly0.7 https://korshunov.ai/en/article/6950-naturebench-evaluates-ai-coding-agents-scientific-discovery-capabilities/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6950-naturebench-otsenivaet-sposobnost-ii-agentov-k-kodirovaniiu-k-nauchnym/2026-06-24weekly0.7 https://korshunov.ai/en/article/6948-adversabench-automated-llm-red-teaming-with-multi-judge-confirmation/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6948-adversabench-avtomatizirovannaia-proverka-uiazvimostei-bolshikh-iazykovykh-s/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6946-metodologicheskaia-ramka-otsenki-sotsialnoi-predvziatosti-v-llm/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6947-memprobe-benchmark-dlia-vosstanovleniia-dolgosrochnoi-pamiati-v-agentakh/2026-06-24weekly0.7 https://korshunov.ai/en/article/6947-memprobe-benchmark-for-long-term-memory-recovery-in-agents/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6945-qwen-agentworld-iazykovye-modeli-mira-dlia-obshchikh-agentov/2026-06-24weekly0.7 https://korshunov.ai/en/article/6945-qwen-agentworld-language-world-models-for-general-agents/2026-06-24weekly0.7 https://korshunov.ai/en/article/6946-methodological-framework-for-evaluating-social-bias-in-llms/2026-06-24weekly0.7 https://korshunov.ai/en/article/6942-sift-and-wsp-improve-fact-checking-accuracy/2026-06-24weekly0.7 https://korshunov.ai/en/article/6943-privacy-preserving-rag-via-multi-agent-semantic-rewriting/2026-06-24weekly0.7 https://korshunov.ai/en/article/6944-cross-lingual-proverb-studies-reveal-cultural-meaning-preservation-in-llms/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6944-issledovaniia-mezhiazychnykh-poslovits-vyiavili-sokhranenie-kulturnogo-v/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6943-anonimizatsiia-rag-za-schet-semanticheskoi-perepiski-mnogoagentnoi-sistemy/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6942-sift-i-wsp-povyshaiut-tochnost-proverki-faktov/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6290-pamiat-effektivnaia-filtratsiia-grafa-dlia-masshtabiruemogo-kollaborativnogo/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6289-szhatie-transformerov-v-rekurrentnye-transformery-dlia-effektivnogo-pamiati/2026-06-24weekly0.7 https://korshunov.ai/en/article/6289-distilling-transformers-into-recurrent-transformers-for-efficient-memory/2026-06-24weekly0.7 https://korshunov.ai/en/article/6290-memory-efficient-graph-filtering-for-scalable-collaborative-filtering/2026-06-24weekly0.7 https://korshunov.ai/en/article/6287-cost-geometry-of-belief-in-noisy-inference/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6288-lig-metody-integrirovannykh-gradientov-na-urovne-sloev-dlia-analiza-potokov-v/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6287-geometriia-stoimosti-very-v-shumnoi-inferentsii/2026-06-24weekly0.7 https://korshunov.ai/en/article/6288-lig-layer-wise-integrated-gradients-for-transformer-flow-analysis/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6286-fast-freimvork-dlia-sinkhronizirovannogo-otbora-i-obucheniia-v-parallelnom/2026-06-24weekly0.7 https://korshunov.ai/en/article/6286-fast-a-framework-for-aligned-sampling-and-training-in-parallel-reinforcement/2026-06-24weekly0.7 https://korshunov.ai/en/article/6285-medlayxplain-benchmarking-expert-lay-gap-in-medical-vision-language-models/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6285-medlayxplain-otsenka-razryva-mezhdu-ekspertami-i-obychnymi-liudmi-v-modeliakh/2026-06-24weekly0.7 https://korshunov.ai/en/article/6284-profile-based-reference-in-llm-grounding/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6284-profilnaia-ssylka-v-fundamentalizatsii-llm/2026-06-24weekly0.7 https://korshunov.ai/en/article/6283-extraction-and-analysis-of-multimodal-concepts-in-vision-language-models/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6283-izvlechenie-i-analiz-multimodalnykh-kontseptsii-v-modeliakh-vizualno-iazykovykh/2026-06-24weekly0.7 https://korshunov.ai/en/article/6282-machine-whistleblowing-a-normative-and-principled-approach/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6282-mashinnoe-raskrytie-informatsii-normativnyi-i-printsipialnyi-podkhod/2026-06-24weekly0.7 https://korshunov.ai/en/article/6281-linguistic-distance-affects-consensus-in-neural-cellular-automata/2026-06-24weekly0.7 https://korshunov.ai/en/article/6937-crewai-1-14-8a3-release-notes/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6280-v-koherence-illiuziiakh-gollandskikh-llms-raskryvaiutsia/2026-06-24weekly0.7 https://korshunov.ai/en/article/6280-coherence-illusions-in-dutch-llms-revealed/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6937-sluzhba-crewai-1-14-8a3-zametki-po-vypusku/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6281-iazykovoe-rasstoianie-vliiaet-na-dostizhenie-soglasiia-v-neironnykh-kletochnykh/2026-06-24weekly0.7 https://korshunov.ai/en/article/6277-fleetagent-efficient-teleoperation-for-autonomous-fleets/2026-06-24weekly0.7 https://korshunov.ai/en/article/6278-improving-speaker-verification-for-non-verbal-vocalizations/2026-06-24weekly0.7 https://korshunov.ai/en/article/6279-qbiofusion-qsar-quantum-kernel-learning-for-small-data-ligand-classification/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6278-uluchshenie-identifikatsii-rechevykh-signalov-dlia-neverbalnykh-zvukov/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6279-qbiofusion-qsar-kvantovoe-iadro-obucheniia-dlia-klassifikatsii-ligandov-na-malom/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6277-fleetagent-effektivnoe-telemanipulirovanie-dlia-avtonomnykh-flotov/2026-06-24weekly0.7 https://korshunov.ai/en/article/6276-recency-frequency-adaptive-kv-caching-for-llm-serving/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6274-empiricheskoe-issledovanie-kvantovaniia-openpangu-na-npus-ascend/2026-06-24weekly0.7 https://korshunov.ai/en/article/6275-ace-gs-efficient-and-accurate-3d-gaussian-splatting/2026-06-24weekly0.7 https://korshunov.ai/en/article/6274-empirical-study-of-openpangu-quantization-on-ascend-npus/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6276-adaptivnoe-keshirovanie-kv-na-osnove-novizny-i-chastoty-ispolzovaniia-dlia/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6275-ace-gs-effektivnaia-i-tochnaia-3d-gaussovaia-splata/2026-06-24weekly0.7 https://korshunov.ai/en/article/6273-arco-adaptive-rubric-with-co-evolution-for-multi-step-llm-agents/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6273-arco-adaptivnyi-rubrik-s-kogeneratsiei-dlia-agentov-na-osnove-mnogoshagovykh-llm/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6272-fastgan-i-modeli-na-osnove-transformerov-uluchshaiut-obnaruzhenie-vreditelei-na/2026-06-24weekly0.7 https://korshunov.ai/en/article/6272-fastgan-and-transformer-models-improve-aphid-detection-in-faba-beans/2026-06-24weekly0.7 https://korshunov.ai/en/article/6270-nasdaq-normalized-observation-space-dynamics-augmented-q-learning/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6270-nasdaq-usrednennaia-dinamika-nabliudenii-s-usileniem-q-obucheniia/2026-06-24weekly0.7 https://korshunov.ai/en/article/6269-influence-based-explanations-for-dysarthria-severity-assessment/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6269-obiasneniia-na-osnove-vliianiia-dlia-otsenki-stepeni-disartirii/2026-06-24weekly0.7 https://korshunov.ai/en/article/6271-topological-neural-dynamics-neuron-wise-sequence-modeling/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6271-topologicheskie-neironnye-dinamiki-modelirovanie-posledovatelnostei-po-neironam/2026-06-24weekly0.7 https://korshunov.ai/en/article/6268-taser-task-differentiated-skill-expansion-for-heterogeneous-continual-learning/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6268-taser-rasshirenie-navykov-na-osnove-zadach-dlia-nepreryvnogo-obucheniia-v/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6265-remont-grafa-ramanuia-umenshaet-perenasyshchenie-v-gnn/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6266-preduprezhdaiushchie-metki-izmeniaiut-vospriiatie-no-ne-vliiaiut-na-vliianie-ii/2026-06-24weekly0.7 https://korshunov.ai/en/article/6266-warning-labels-shift-perceptions-but-not-ai-influence-of-sycophancy/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6267-model-sotsialnogo-mira-dlia-dolgosrochnoi-sotsialnoi-intellektualnosti/2026-06-24weekly0.7 https://korshunov.ai/en/article/6267-social-world-model-for-lifelong-social-intelligence/2026-06-24weekly0.7 https://korshunov.ai/en/article/6265-ramanujan-graph-rewiring-alleviates-gnn-over-squashing/2026-06-24weekly0.7 https://korshunov.ai/en/article/6262-sohet-self-supervised-transformer-for-heterogeneous-event-streams/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6262-sohet-samosupervizirovannyi-transformator-dlia-geterogennykh-potokov-sobytii/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6264-dataclaw0-agentnaia-nastroika-multimodalnykh-dannykh-iz-iskhodnykh-potokov/2026-06-24weekly0.7 https://korshunov.ai/en/article/6263-transformer-models-highly-sensitive-to-noisy-data-in-trajectory-prediction/2026-06-24weekly0.7 https://korshunov.ai/en/article/6264-dataclaw0-agentic-tailoring-of-multimodal-data-from-raw-streams/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6263-modeli-transformer-chrezvychaino-chuvstvitelny-k-shumu-v-dannykh-o-traektorii/2026-06-24weekly0.7 https://korshunov.ai/en/article/6822-mimo-2-5-is-fast-at-large-context-on-dual-rtx-pro-6000/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6822-mimo-2-5-rabotaet-bystro-pri-bolshikh-kontekstakh-na-dvukh-kartakh-rtx-pro-6000/2026-06-24weekly0.7 https://korshunov.ai/ru/article/6815-datasette-1-0a35-vypuskaet-novye-funktsii-sozdaniia-i-izmeneniia-tablits/2026-06-23weekly0.7 https://korshunov.ai/en/article/6815-datasette-1-0a35-releases-new-table-creation-and-alteration-features/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6260-seksualizirovannye-ii-golosa-usilivaiut-gendernye-nesootvetstviia-v-voprosakh/2026-06-23weekly0.7 https://korshunov.ai/en/article/6260-sexualised-ai-voices-amplify-gender-power-asymmetries/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6261-lambdamark-pervyi-obshchii-skhema-radioaktivnoi-audio-vstavki/2026-06-23weekly0.7 https://korshunov.ai/en/article/6261-lambdamark-first-generic-radioactive-audio-watermarking-scheme/2026-06-23weekly0.7 https://korshunov.ai/en/article/6259-graph-of-differences-for-anatomy-structured-medreid/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6259-graf-raznostei-dlia-identifikatsii-meditsinskikh-izobrazhenii-s-uchetom/2026-06-23weekly0.7 https://korshunov.ai/en/article/6256-llms-benchmarked-for-web-vulnerability-detection/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6256-otsenka-llm-dlia-obnaruzheniia-uiazvimostei-v-veb-prilozheniiakh/2026-06-23weekly0.7 https://korshunov.ai/en/article/6257-functional-orthogonality-ensures-identifiability-in-unsupervised-disentanglement/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6257-funktsionalnaia-ortogonalnost-obespechivaet-identifitsiruemost-v-neperemennoi/2026-06-23weekly0.7 https://korshunov.ai/en/article/6258-entrust-modeling-inter-modal-conflict-for-trustworthy-multimodal-medical-image/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6258-entrust-modelirovanie-mezhmodalnykh-konfliktov-dlia-nadezhnogo-analiza/2026-06-23weekly0.7 https://korshunov.ai/en/article/6254-swarmx-agentic-scheduling-for-low-latency-systems/2026-06-23weekly0.7 https://korshunov.ai/en/article/6255-llm-agent-oversight-must-shift-from-calibration-to-action-conditioned-control/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6252-2d-protiv-3d-diffuzii-dlia-sinteticheskoi-podgotovki-ii-na-rentgenovskikh/2026-06-23weekly0.7 https://korshunov.ai/en/article/6252-2d-vs-3d-diffusion-for-synthetic-x-ray-ai-training/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6255-oversait-llm-agenta-dolzhen-pereiti-ot-kalibrovki-k-kontroliu-zavisiashchemu-ot/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6254-swarmx-agentskoe-planirovanie-dlia-sistem-s-nizkoi-zaderzhkoi/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6253-nadezhnaia-obratnaia-sviaz-mozhet-nanesti-vred-agentam-s-ispolzovaniem/2026-06-23weekly0.7 https://korshunov.ai/en/article/6253-unreliable-feedback-can-harm-tool-using-llm-agents/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6251-mircaps-masshtabnyi-mnogodomennyi-vizualno-iazykovoi-dataset/2026-06-23weekly0.7 https://korshunov.ai/en/article/6251-mircaps-large-scale-mixed-domain-vision-language-dataset/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6814-650-modelei-ner-dezidentifikatsii-na-osnove-biomeditsinskikh-dannykh-pod-apache/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6250-modeli-moe-pokazyvaiut-zavisimost-proizvoditelnosti-ot-ustroistva/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6249-autoras-obuchenie-ustoichivykh-agentskikh-sistem-s-primitivnymi-predstavleniiami/2026-06-23weekly0.7 https://korshunov.ai/en/article/6814-650-apache-2-0-biomedical-ner-de-id-models-run-30-40x-faster-on-apple-silicon/2026-06-23weekly0.7 https://korshunov.ai/en/article/6250-moe-models-show-device-dependent-inference-performance/2026-06-23weekly0.7 https://korshunov.ai/en/article/6249-autoras-learning-robust-agentic-systems-with-primitive-representations/2026-06-23weekly0.7 https://korshunov.ai/en/article/6247-benchmark-evaluation-of-small-language-models-for-arabic-nlp/2026-06-23weekly0.7 https://korshunov.ai/en/article/6246-explainable-ai-model-for-career-related-depression-in-university-students/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6246-model-obiasnimoi-ii-dlia-depressii-sviazannoi-s-kareroi-u-studentov/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6247-otsenka-benchmarka-malykh-iazykovykh-modelei-dlia-arabskoi-nlp/2026-06-23weekly0.7 https://korshunov.ai/en/article/6248-cortis-text-only-adaptation-of-spoken-language-models/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6248-cortis-tekstovoe-adaptatsiia-modelei-rechi/2026-06-23weekly0.7 https://korshunov.ai/en/article/6241-analytic-policy-gradients-for-sample-and-learning-efficient-control/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6241-analiticheskie-gradienty-politik-dlia-effektivnogo-upravleniia-obraztsami-i/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6245-mashinnoe-obuchenie-predskazyvaet-vysokoriskovye-polipy-kishechnika-u/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6244-razdelenie-deklarativnykh-i-protsedurnykh-znanii-v-modeliakh-vizualno-iazykovykh/2026-06-23weekly0.7 https://korshunov.ai/en/article/6245-machine-learning-model-predicts-high-risk-colorectal-polyps-in-african-americans/2026-06-23weekly0.7 https://korshunov.ai/en/article/6244-decoupling-declarative-and-procedural-knowledge-in-vision-language-action-models/2026-06-23weekly0.7 https://korshunov.ai/en/article/6242-js-divergence-improves-grpo-autoregressive-text-to-image-alignment/2026-06-23weekly0.7 https://korshunov.ai/en/article/6243-deep-learning-with-o-log-n-parallel-time-complexity/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6243-glubokoe-obuchenie-s-parallelnoi-vremennoi-slozhnostiu-o-log-n/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6242-js-razdelenie-uluchshaet-autoregressivnuiu-alignment-teksta-izobrazheniia-v-grpo/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6238-proverka-na-etape-proektirovaniia-agentnykh-ii-protsessov/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6240-pamiat-effektivnaia-filtratsiia-grafa-dlia-masshtabiruemogo-kratkosrochnogo/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6239-sotsialnaia-teoriia-vybora-v-oblasti-vyravnivaniia-ii/2026-06-23weekly0.7 https://korshunov.ai/en/article/6239-ai-alignment-via-social-choice-theory/2026-06-23weekly0.7 https://korshunov.ai/en/article/6240-memory-efficient-graph-filtering-for-scalable-collaborative-filtering/2026-06-23weekly0.7 https://korshunov.ai/en/article/6238-design-time-verification-of-agentic-ai-workflows/2026-06-23weekly0.7 https://korshunov.ai/en/article/6237-zero-shot-procedural-mistake-detection-with-vlms/2026-06-23weekly0.7 https://korshunov.ai/en/article/6807-adding-second-7900xtx-via-pcie-2-on-b450-motherboard/2026-06-23weekly0.7 https://korshunov.ai/en/article/6236-fast-a-framework-for-aligned-sampling-and-training-in-parallel-reinforcement/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6237-obnaruzhenie-oshibok-v-protsedure-bez-obucheniia-s-ispolzovaniem-modelei-video/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6807-dobavlenie-vtorogo-7900xtx-cherez-pcie-2-na-materinskoi-plate-b450/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6236-fast-freimvork-dlia-sinkhronizirovannogo-otbora-i-obucheniia-v-parallelnom/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6799-kakie-samye-populiarnye-platformy-arendy-gpu-v-kitae/2026-06-23weekly0.7 https://korshunov.ai/en/article/6798-minimax-2-7-runs-on-47tg-1200pp-with-96gb-vram/2026-06-23weekly0.7 https://korshunov.ai/en/article/6799-what-are-the-top-chinese-gpu-rental-platforms/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6798-minimax-2-7-rabotaet-na-47-tg-1200pp-s-96-gb-vram/2026-06-23weekly0.7 https://korshunov.ai/en/article/6800-claude-v2-1-187-release-notes/2026-06-23weekly0.7 https://korshunov.ai/en/article/6794-open-source-hugging-face-downloader-app-released/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6800-primechaniia-po-vypusku-claude-v2-1-187/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6794-vypushchen-otkrytyi-iskhodnyi-prilozhenie-hugging-face-dlia-zagruzki/2026-06-23weekly0.7 https://korshunov.ai/en/article/6791-tmax-27b-terminal-agent-for-small-gpus-with-dppo-training/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6791-agent-tmax-27b-dlia-malykh-videokart-s-obucheniem-dppo/2026-06-23weekly0.7 https://korshunov.ai/en/article/6780-openmythos-benchmarks-released-with-swe-bench-and-cybersecurity-results/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6780-opublikovany-benchmarki-openmythos-s-rezultatami-swe-bench-i-v-oblasti/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6779-opublikovany-benchmarki-openmythos-s-rezultatami-swe-bench-i-v-oblasti/2026-06-23weekly0.7 https://korshunov.ai/en/article/6779-openmythos-benchmarks-released-with-swe-bench-and-cybersecurity-results/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6793-openai-razvivaet-obshchie-standarty-ii-cherez-appia-foundation/2026-06-23weekly0.7 https://korshunov.ai/en/article/6793-openai-builds-shared-ai-standards-via-appia-foundation/2026-06-23weekly0.7 https://korshunov.ai/en/article/6773-nvidia-claims-15x-speedup-with-diffusion-model/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6774-opfs-pyodide-test-harness-dlia-redaktirovaniia-sqlite-v-brauzere/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6773-nvidia-utverzhdaet-15-kratnoe-uskorenie-pri-ispolzovanii-modeli-diffuzii/2026-06-23weekly0.7 https://korshunov.ai/en/article/6774-opfs-pyodide-test-harness-for-browser-based-sqlite-editing/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6786-dostup-k-api-llama-3-1-70b-ogranichen-dlia-testerov-hugging-face/2026-06-23weekly0.7 https://korshunov.ai/en/article/6785-native-binary-embeddings-outperform-post-hoc-binarization/2026-06-23weekly0.7 https://korshunov.ai/en/article/6786-llama-3-1-70b-api-access-restricted-to-hugging-face-tester/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6785-nativnye-binarnye-vektory-prevoskhodiat-post-posledovatelnuiu-binarizatsiiu/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6787-informatsiia-o-postavshchike-inferensa-ustarela/2026-06-23weekly0.7 https://korshunov.ai/en/article/6787-inference-provider-information-out-of-date/2026-06-23weekly0.7 https://korshunov.ai/en/article/6783-seeking-arxiv-cs-lg-endorsement-for-psilogic-optimizer/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6782-tokeny-spaces-bolshe-ne-rabotaiut-i-faily-ne-sokhraniaiutsia/2026-06-23weekly0.7 https://korshunov.ai/en/article/6782-spaces-tokens-no-longer-work-and-files-not-saved/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6783-prosba-o-podderzhke-arxiv-cs-lg-dlia-optimizatora-psilogic/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6784-tokeny-spaces-perestali-rabotat-posle-obnovleniia/2026-06-23weekly0.7 https://korshunov.ai/en/article/6784-spaces-tokens-stop-working-after-update/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6769-novye-qwen-27b-iq4-ks-i-iq4-ks-kt-dlia-ik-llama-cpp/2026-06-23weekly0.7 https://korshunov.ai/en/article/6769-new-qwen-27b-iq4-ks-and-iq4-ks-kt-quantizations-for-ik-llama-cpp/2026-06-23weekly0.7 https://korshunov.ai/en/article/6763-can-glm5-2-be-run-on-4x-amd-epyc-servers-with-512gb-ram-each/2026-06-23weekly0.7 https://korshunov.ai/en/article/6762-not-ironclad-confirmation-but/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6762-ne-okonchatelnoe-podtverzhdenie-no/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6763-mozhno-li-zapustit-glm5-2-na-4-serverakh-amd-epyc-s-512-gb-ozu-kazhdyi/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6750-tseny-na-modeli-openrouter-ukazyvaiut-na-bolee-glubokuiu-kvantovanie/2026-06-23weekly0.7 https://korshunov.ai/en/article/6750-openrouter-model-prices-imply-heavier-quantization/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6749-glm-5-2-na-mac-studio-uskorenie-pr/2026-06-23weekly0.7 https://korshunov.ai/en/article/6749-glm-5-2-on-mac-studio-speedup-pr/2026-06-23weekly0.7 https://korshunov.ai/en/article/6744-llm-medical-scribing-benchmark-omissions-outnumber-hallucinations/2026-06-23weekly0.7 https://korshunov.ai/en/article/6743-7-chinese-companies-shipping-h100-h200-class-ai-chips-most-ipo-d-in-last-6/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6743-7-kitaiskikh-kompanii-postavliaiut-chipy-klassa-h100-h200-bolshinstvo-iz-byli-v/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6744-benchmark-llm-v-meditsinskoi-sintaksise-propuski-prevoskhodiat-vydumki/2026-06-23weekly0.7 https://korshunov.ai/en/article/6737-vibethinker-3b-parameter-model-beats-opus-4-5-in-reasoning/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6737-vibethinker-model-s-3-milliardami-parametrov-pobezhdaet-opus-4-5-v-zadachakh/2026-06-23weekly0.7 https://korshunov.ai/en/article/6735-krea-2-released-on-hugging-face/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6736-analiz-kld-kvantovaniia-kesha-kv-dlia-modelei-qwen3-6-35b-a3b-i-gemma4-e2b-qat/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6735-krea-2-vypushchen-na-hugging-face/2026-06-23weekly0.7 https://korshunov.ai/en/article/6736-kld-analysis-of-kv-cache-quantization-for-qwen3-6-35b-a3b-and-gemma4-e2b-qat/2026-06-23weekly0.7 https://korshunov.ai/en/article/6764-frontier-llms-struggle-to-write-fast-multi-gpu-kernels/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6764-frontier-llms-struggle-to-write-fast-multi-gpu-kernels/2026-06-23weekly0.7 https://korshunov.ai/en/article/6765-testing-cross-origin-storage-api-in-transformers-js/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6765-proverka-api-khraneniia-dannykh-mezhdu-domenami-v-transformers-js/2026-06-23weekly0.7 https://korshunov.ai/en/article/6731-baidu-releases-one-shot-long-horizon-parsing/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6731-baidu-vypustil-model-odnoshagovogo-dolgosrochnogo-parsinga/2026-06-23weekly0.7 https://korshunov.ai/en/article/6716-comparing-docling-liteparse-mineru-and-unstructured-for-on-prem-document/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6716-sravnenie-docling-liteparse-mineru-i-unstructured-dlia-obrabotki-dokumentov-na/2026-06-23weekly0.7 https://korshunov.ai/en/article/6713-my-new-benchmark-how-good-are-llms-at-simulating-wetting-behavior/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6758-vvedenie-claude-tag-dlia-komand-slack/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6713-moi-novyi-benchmark-naskolko-khorosho-llm-mogut-imitirovat-povedenie/2026-06-23weekly0.7 https://korshunov.ai/en/article/6758-introducing-claude-tag-for-slack-teams/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6714-llama-server-webui-ne-otvechaet-posle-perekompiliatsii/2026-06-23weekly0.7 https://korshunov.ai/en/article/6714-llama-server-webui-not-responding-after-recompile/2026-06-23weekly0.7 https://korshunov.ai/en/article/6703-cpu-only-tts-benchmark-kokoro-82m-vs-supertonic-3-vs-inflect-nano-v1/2026-06-23weekly0.7 https://korshunov.ai/en/article/6701-pooled-round-robin-hardware-with-friends/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6701-sobiranie-vrashchaiushchegosia-kolesa-apparatury-s-druziami/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6703-otsenka-tts-bez-ispolzovaniia-cpu-kokoro-82m-protiv-supertonic-3-protiv-inflect/2026-06-23weekly0.7 https://korshunov.ai/en/article/6752-vulkan-backend-updates-and-new-binary-releases-for-llama-cpp/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6752-obnovleniia-bekenda-vulkan-i-novye-versii-binarnykh-failov-dlia-llama-cpp/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6748-gpt-5-pro-pomog-reshit-zagadku-immunologii-sushchestvovavshuiu-tri-goda/2026-06-23weekly0.7 https://korshunov.ai/en/article/6751-llama-cpp-release-b9775-new-binaries-and-support-for-multiple-platforms/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6751-reliz-llama-cpp-b9775-novye-binarniki-i-podderzhka-neskolkikh-platform/2026-06-23weekly0.7 https://korshunov.ai/en/article/6748-gpt-5-pro-helps-solve-3-year-old-immunology-mystery/2026-06-23weekly0.7 https://korshunov.ai/en/article/6690-reusable-workflows-for-long-running-local-llms/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6690-ispolzuemye-rabochie-protsessy-dlia-dlitelnykh-lokalnykh-modelei-iazykovoi/2026-06-23weekly0.7 https://korshunov.ai/en/article/6688-v100-4-card-ai-large-model-with-tesla-128g-server/2026-06-23weekly0.7 https://korshunov.ai/en/article/6689-review-of-jackrong-qwopus3-5-9b-coder-mtp-gguf/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6688-model-ai-s-chetyrmia-kartochkami-v100-v-servere-tesla-128-gb/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6689-obzor-jackrong-qwopus3-5-9b-coder-mtp-gguf/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6683-pozitsiia-glm-5-2-otrazhaet-vliianie-kulturnoi-podgotovki/2026-06-23weekly0.7 https://korshunov.ai/en/article/6683-glm-5-2-s-attitude-reflects-cultural-training-influences/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6681-moi-lokalnyi-server-prostoial-99-vremeni/2026-06-23weekly0.7 https://korshunov.ai/en/article/6681-my-local-server-idling-99-of-the-time/2026-06-23weekly0.7 https://korshunov.ai/en/article/6738-llama-cpp-release-b9773-adds-vulkan-support-and-new-builds/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6738-llama-cpp-release-b9773-dobavliaet-podderzhku-vulkan-i-novye-sborki/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6672-luchshie-lokalnye-modeli-dlia-rassuzhdenii-v-agentnoi-ii/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6671-chelovecheskaia-otsenka-pokazyvaet-chto-glm-5-2-konkuriruet-s-luchshimi/2026-06-23weekly0.7 https://korshunov.ai/en/article/6671-human-evaluation-shows-glm-5-2-competes-with-top-models/2026-06-23weekly0.7 https://korshunov.ai/en/article/6672-best-local-models-for-reasoning-in-agentic-ai/2026-06-23weekly0.7 https://korshunov.ai/en/article/6669-multi-tier-moe-caching-optimizing-expert-activation-in-large-models/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6669-mnogourovnevoe-keshirovanie-moe-optimizatsiia-aktivatsii-ekspertov-v-krupnykh/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6660-kak-dokazat-chto-lokalnaia-model-llm-ne-sobiraet-dannye-polzovatelei/2026-06-23weekly0.7 https://korshunov.ai/en/article/6660-how-to-prove-an-llm-app-doesn-t-collect-user-data/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6658-sft-ili-rl-pervyi-dlia-obucheniia-agenta-qwen-3-5-s-ispolzovaniem-instrumentov/2026-06-23weekly0.7 https://korshunov.ai/en/article/6658-sft-or-rl-first-for-qwen-3-5-tool-agent-training/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6653-pochemu-gemma-4-26b-ne-upominaetsia-chashche/2026-06-23weekly0.7 https://korshunov.ai/en/article/6653-why-is-gemma-4-26b-not-mentioned-more/2026-06-23weekly0.7 https://korshunov.ai/en/article/6650-predicate-importance-estimation-and-decoupled-rationale-score-distillation-for/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6649-gruppovaia-grafovaia-optimizatsiia-politiki-dlia-dolgosrochnoi-agentnoi-rl/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6650-otsenka-vazhnosti-predikatov-i-dekuplirovannaia-distilliatsiia-obosnovanii-dlia/2026-06-23weekly0.7 https://korshunov.ai/en/article/6649-group-graph-policy-optimization-for-long-horizon-agentic-rl/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6721-reliz-llama-cpp-b9771-dobavliaet-podderzhku-vulkan-i-optimizatsii/2026-06-23weekly0.7 https://korshunov.ai/en/article/6723-mistral-releases-ocr-4-with-multilingual-support-and-structured-output/2026-06-23weekly0.7 https://korshunov.ai/en/article/6721-llama-cpp-release-b9771-adds-vulkan-support-and-optimizations/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6723-mistral-vypustil-ocr-4-s-podderzhkoi-neskolkikh-iazykov-i-strukturirovannym/2026-06-23weekly0.7 https://korshunov.ai/en/article/6646-language-model-panel-for-political-position-measurement-in-data-sparse-regions/2026-06-23weekly0.7 https://korshunov.ai/en/article/6648-comparative-evaluation-of-mt-systems-and-post-editor-groups-in-specialised/2026-06-23weekly0.7 https://korshunov.ai/en/article/6647-entity-level-membership-inference-via-llm-interrogation/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6647-uroven-sushchnosti-v-vyiavlenii-chlenstva-cherez-interviu-s-llm/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6648-sravnitelnaia-otsenka-sistem-mashinnogo-perevoda-i-grupp-postredaktorov-v/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6646-panel-iazykovykh-modelei-dlia-izmereniia-politicheskikh-pozitsii-v-regionakh-s/2026-06-23weekly0.7 https://korshunov.ai/en/article/6645-phonebuddy-training-open-models-for-agentic-phone-use/2026-06-23weekly0.7 https://korshunov.ai/en/article/6644-unlimited-ocr-human-like-parsing-with-constant-memory/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6645-phonebuddy-obuchenie-otkrytykh-modelei-dlia-agentnogo-ispolzovaniia-telefona/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6644-neogranichennyi-ocr-parsing-pokhozhii-na-chelovecheskii-s-postoiannym-pamiati/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6642-pivotsbench-benchmark-dlia-detalnogo-rassuzhdeniia-o-mezhlichnostnykh-v-mllm/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6643-vladenie-rekomendatsiiami-ii-empiricheskaia-karta-vladeniia-brendami-v-kategorii/2026-06-23weekly0.7 https://korshunov.ai/en/article/6643-ai-recommendation-ownership-empirical-map-of-brand-category-ownership/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6641-kognitivnye-tsifrovye-dvoiniki-eticheskie-riski-i-upravlenie/2026-06-23weekly0.7 https://korshunov.ai/en/article/6642-pivotsbench-benchmark-for-fine-grained-interpersonal-reasoning-in-mllms/2026-06-23weekly0.7 https://korshunov.ai/en/article/6641-cognitive-digital-twins-ethical-risks-and-governance/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6636-koshur-pixel-pervyi-masshtabnyi-sinteticheskii-nabor-dannykh-dlia-raspoznavaniia/2026-06-23weekly0.7 https://korshunov.ai/en/article/6636-koshur-pixel-first-large-scale-synthetic-ocr-dataset-for-kashmiri/2026-06-23weekly0.7 https://korshunov.ai/en/article/6640-dual-track-framework-for-template-constrained-latex-conversion/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6639-samorazvitie-agentov-ispolzuiushchikh-instrumenty-cherez-obuchenie-predpochtenii/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6640-dvukhkanalnyi-freimvork-dlia-preobrazovaniia-latex-s-uchetom-shablonov/2026-06-23weekly0.7 https://korshunov.ai/en/article/6639-self-evolution-of-tool-calling-agents-via-divergence-point-preference-learning/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6638-pride-metod-rasshireniia-znanii-s-ispolzovaniem-privilegirovannykh-dannykh-dlia/2026-06-23weekly0.7 https://korshunov.ai/en/article/6638-pride-privileged-information-enhanced-distillation-for-empathetic-dialogue/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6637-after-otsenivaet-pamiat-o-protsedurakh-v-agentakh-na-osnove-llm/2026-06-23weekly0.7 https://korshunov.ai/en/article/6637-after-benchmark-evaluates-procedural-memory-in-llm-agents/2026-06-23weekly0.7 https://korshunov.ai/en/article/6708-coolest-theoretical-ai-topics-with-realistic-ai-system-basis/2026-06-23weekly0.7 https://korshunov.ai/en/article/6707-apakah-layanan-wa-bank-bni-24-jam-ini-nomor-call-center/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6708-samye-interesnye-teoreticheskie-temy-ii-s-realnoi-bazoi-ii-sistem/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6707-rabotaet-li-servis-wa-bank-bni-24-chasa-nomer-tsentra-podderzhki/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6705-reliz-llama-cpp-b9770-ispravlenie-obrabotki-udalennykh-shablonov-i-dobavlenie/2026-06-23weekly0.7 https://korshunov.ai/en/article/6705-llama-cpp-release-b9770-fixes-remote-preset-handling-and-adds-tests/2026-06-23weekly0.7 https://korshunov.ai/en/article/6704-build-real-agentic-apps-with-cuga-24-working-examples/2026-06-23weekly0.7 https://korshunov.ai/en/article/6711-ai-s-cultural-gaps-expose-global-users-to-misrepresentation-and-marginalization/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6711-kulturnye-probely-v-ii-obnazhaiut-globalnykh-polzovatelei-pered-iskazheniem-i/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6704-sozdanie-realnykh-agentnykh-prilozhenii-s-cuga-24-rabochikh-primera/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6635-iazyk-formiruet-istoricheskuiu-kreditnost-v-bolshikh-iazykovykh-modeliakh/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6633-model-tts-na-osnove-potoka-sootvetstviia-imitiruet-effekt-lombarda/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6634-ai-sozdannaia-reputatsiia-brenda-zavisit-ot-iazyka/2026-06-23weekly0.7 https://korshunov.ai/en/article/6635-language-shapes-historical-credit-in-large-language-models/2026-06-23weekly0.7 https://korshunov.ai/en/article/6633-flow-matching-tts-model-simulates-lombard-effect/2026-06-23weekly0.7 https://korshunov.ai/en/article/6634-ai-constructed-brand-reputation-is-language-bound/2026-06-23weekly0.7 https://korshunov.ai/en/article/6631-agentcibench-evaluates-privacy-risks-in-computer-use-agents/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6630-kontagioznoe-rasprostranenie-pamiati-rasprostranenie-predvziatosti-v-pamiati/2026-06-23weekly0.7 https://korshunov.ai/en/article/6632-dart-training-free-routing-for-adaptive-thinking-budgets/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6632-dart-obuchenie-besplatnoe-perenapravlenie-dlia-adaptivnykh-biudzhetov-myshleniia/2026-06-23weekly0.7 https://korshunov.ai/en/article/6630-memory-contagion-bias-propagation-in-agent-memory/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6631-agentcibench-otsenivaet-riski-privatnosti-v-agentakh-ispolzuiushchikh-kompiuter/2026-06-23weekly0.7 https://korshunov.ai/en/article/6629-task-sensitive-analysis-of-intrinsic-self-correction/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6629-analiz-zadachospetsifichnosti-vrozhdennoi-samokorrektsii/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6628-cfpo-counterfactual-policy-optimization-dlia-mnogomodalnogo-myshleniia/2026-06-23weekly0.7 https://korshunov.ai/en/article/6628-cfpo-counterfactual-policy-optimization-for-multimodal-reasoning/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6627-muppet-benchmark-dlia-privatnosti-mnogopartiinykh-llm/2026-06-23weekly0.7 https://korshunov.ai/en/article/6627-muppet-benchmark-for-multi-party-llm-privacy/2026-06-23weekly0.7 https://korshunov.ai/en/article/6623-segmentation-width-and-cluster-size-impact-speech-resynthesis-in-gslms/2026-06-23weekly0.7 https://korshunov.ai/en/article/6625-kdos-distribution-optimized-synthesis-for-llm-knowledge-expansion/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6626-rasshirenie-osnovannoe-na-otsenke-dlia-generatsii-ekspertnoi-otsenki/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6625-kdos-raspredelenie-optimizirovannoe-sintezirovanie-dlia-rasshireniia-znanii-v/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6623-shirina-segmentatsii-i-razmer-klastera-vliiaiut-na-rezinsis-rechi-v-modeliakh/2026-06-23weekly0.7 https://korshunov.ai/en/article/6626-judgment-grounded-expansion-for-peer-review-generation/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6624-imlogic-benchmark-i-rootmem-framework-dlia-neiavnogo-logicheskogo-pamiati/2026-06-23weekly0.7 https://korshunov.ai/en/article/6624-imlogic-benchmark-and-rootmem-framework-for-implicit-logical-memory-retrieval/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6622-razryv-orakulnoi-ctc-akusticheskaia-vygoranie-i-lingvisticheskaia/2026-06-23weekly0.7 https://korshunov.ai/en/article/6622-ctc-oracle-gap-acoustic-exhaustion-and-linguistic-recovery/2026-06-23weekly0.7 https://korshunov.ai/en/article/6694-granite-speech-plus-conversion-support-added/2026-06-23weekly0.7 https://korshunov.ai/en/article/6693-fix-for-vulkan-result-checking-and-test-linking-in-llama-cpp/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6693-ispravlenie-proverki-rezultatov-vulkan-i-sviazi-testov-v-llama-cpp/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6694-dobavleno-podderzhku-preobrazovaniia-granite-speech-plus/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6621-neopredelennost-orientirovannaia-dezinfitsirovka-dlia-dezinfitsirovaniia-llm/2026-06-23weekly0.7 https://korshunov.ai/en/article/6621-uncertainty-based-decontamination-for-llm-decontamination/2026-06-23weekly0.7 https://korshunov.ai/en/article/6620-tmax-a-simple-rl-recipe-for-terminal-agents/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6620-tmax-prostaia-retseptura-rl-dlia-agentov-konechnykh/2026-06-23weekly0.7 https://korshunov.ai/en/article/6619-wavedetect-framework-for-machine-generated-text-detection-via-wavelet-transform/2026-06-23weekly0.7 https://korshunov.ai/en/article/6618-tf-refusalbench-measures-over-alignment-in-llms-for-criminal-law/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6619-wavedetect-freimvork-dlia-obnaruzheniia-mashinno-generiruemogo-teksta-s-volny/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6618-tf-refusalbench-izmeriaet-chrezmernuiu-sinkhronizatsiiu-v-llms-dlia-ugolovnogo/2026-06-23weekly0.7 https://korshunov.ai/en/article/6617-energy-based-transformers-predict-reading-difficulty/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6615-pri-etom-prostranstva-vstraivaniia-llm-vosstanavlivaiut-strukturu-ekspertov/2026-06-23weekly0.7 https://korshunov.ai/en/article/6615-do-llm-embedding-spaces-recover-expert-structure/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6617-transformatory-na-osnove-energii-predskazyvaiut-slozhnost-chteniia/2026-06-23weekly0.7 https://korshunov.ai/en/article/6616-self-stigma-is-not-uniform-llms-need-persona-aware-support/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6616-samostigma-ne-iavliaetsia-odnorodnoi-llms-nuzhdaiutsia-v-podderzhke-personazh/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6612-triggerbench-otsenka-perspektivnoi-pamiati-v-bolshikh-iazykovykh-modeliakh/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6613-unbias-plus-vyiavlenie-obiasnenie-i-perepisyvanie-predvziatosti/2026-06-23weekly0.7 https://korshunov.ai/en/article/6612-triggerbench-evaluating-prospective-memory-in-llms/2026-06-23weekly0.7 https://korshunov.ai/en/article/6614-reasoninglens-hierarchical-visualization-for-large-reasoning-models/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6614-reasoninglens-ierarkhicheskaia-vizualizatsiia-dlia-bolshikh-modelei-rassuzhdenii/2026-06-23weekly0.7 https://korshunov.ai/en/article/6613-unbias-plus-detect-explain-and-rewrite-bias/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6611-rost-voenno-orientirovannoi-leksiki-v-nauchnykh-abstraktakh/2026-06-23weekly0.7 https://korshunov.ai/en/article/6610-selfcompact-self-driving-context-compaction-for-language-models/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6609-verievol-masshtabirovanie-mnogomodalnykh-matematicheskikh-rassuzhdenii-s/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6610-selfcompact-samodvizhushchiisia-szhatie-konteksta-dlia-iazykovykh-modelei/2026-06-23weekly0.7 https://korshunov.ai/en/article/6609-verievol-scaling-multimodal-mathematical-reasoning-with-verifiable-evolution/2026-06-23weekly0.7 https://korshunov.ai/en/article/6611-militarized-language-rising-in-scientific-abstracts/2026-06-23weekly0.7 https://korshunov.ai/en/article/6606-svd-surgeon-optimal-singular-value-surgery-for-llm-compression/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6606-svd-surgeon-optimalnaia-operatsiia-po-znacheniiam-singuliarnykh-dlia-szhatiia/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6608-model-energopotrebleniia-dlia-obucheniia-transformatorov/2026-06-23weekly0.7 https://korshunov.ai/en/article/6608-energy-consumption-model-for-transformer-training/2026-06-23weekly0.7 https://korshunov.ai/en/article/6607-langmap-language-adaptive-tokenization-for-multilingual-models/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6607-langmap-adaptivnaia-tokenizatsiia-na-iazyke-dlia-modelei-na-neskolkikh-iazykakh/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6604-enterpriseclawbench-vypushchena-otsenka-agentov-v-realnykh-usloviiakh/2026-06-23weekly0.7 https://korshunov.ai/en/article/6604-enterpriseclawbench-real-world-agent-benchmark-released/2026-06-23weekly0.7 https://korshunov.ai/en/article/6603-tapered-language-models-improve-performance/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6603-modeli-s-umenshaiushchimsia-kolichestvom-parametrov-povyshaiut-proizvoditelnost/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6605-otsenka-osoznaniia-iavliaetsia-mnogomernoi-a-ne-odnorodnym-svoistvom/2026-06-23weekly0.7 https://korshunov.ai/en/article/6605-evaluation-awareness-is-multivariate-not-a-single-capability/2026-06-23weekly0.7 https://korshunov.ai/en/article/6601-randomized-yarn-improves-length-generalization-for-long-context-reasoning/2026-06-23weekly0.7 https://korshunov.ai/en/article/6602-llms-fail-to-reliably-self-report-adversarial-prefills/2026-06-23weekly0.7 https://korshunov.ai/en/article/6500-is-sakana-fugu-just-an-iq-experiment/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6500-iavliaetsia-li-sakana-fugu-prosto-eksperimentom-po-ik/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6601-sluchainyi-yarn-uluchshaet-obobshchenie-po-dline-dlia-rassuzhdenii-s-dlinnymi/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6602-llms-ne-sposobny-nadezhno-otchetlivo-otrazhat-vliianie-atak-na-predvaritelnye/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6499-model-minimax-m3-eagle3-gguf-teper-sovmestima-s-llama-cpp/2026-06-23weekly0.7 https://korshunov.ai/en/article/6498-100-t-s-on-qwen3-6-27b-q8-across-5090-3090-ti-with-tensor-split-mode/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6498-100-t-s-na-qwen3-6-27b-q8-0-s-ispolzovaniem-5090-i-3090-ti-v-rezhime-tensor-mode/2026-06-23weekly0.7 https://korshunov.ai/en/article/6499-minimax-m3-eagle3-gguf-model-now-compatible-with-llama-cpp/2026-06-23weekly0.7 https://korshunov.ai/en/article/6491-eu-ai-act-mandates-ai-generated-text-watermarking-from-august-2024/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6493-ogranicheniia-graficheskoi-karty-intel-arc-b70/2026-06-23weekly0.7 https://korshunov.ai/en/article/6493-intel-arc-b70-gpu-limitations-exposed/2026-06-23weekly0.7 https://korshunov.ai/en/article/6492-boogu-image-0-1-open-source-unified-image-generation-and-editing-model-series/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6492-boogu-image-0-1-seriia-otkrytykh-istochnikov-universalnykh-modelei-generatsii-i/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6491-zakon-es-o-iskusstvennom-intellekte-trebuet-vodianykh-znakov-teksta-ii-s-avgusta/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6389-kto-nuzhdaetsia-v-gpu-64-t-s-gen-285-pp-na-protsessorakh-vozrastom-6-let/2026-06-23weekly0.7 https://korshunov.ai/en/article/6388-sk-hynix-delays-hbm4-shift-to-boost-dram-production/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6388-sk-hynix-otkladyvaet-perekhod-na-hbm4-dlia-uvelicheniia-proizvodstva-dram/2026-06-23weekly0.7 https://korshunov.ai/en/article/6389-who-needs-gpus-64-t-s-gen-285-pp-on-6-year-old-cpus/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6385-capricorn-1k-benchmark-dlia-opisaniia-video-i-soglasovannosti-temy/2026-06-23weekly0.7 https://korshunov.ai/en/article/6385-capricorn-1k-benchmark-for-video-captioning-and-subject-consistency/2026-06-23weekly0.7 https://korshunov.ai/en/article/6386-symmetric-q-sorts-measure-value-structure-alignment-in-llms/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6384-na-samom-dele-li-mnogoznachnye-modeli-uluchshaiutsia-izoliatsiia-istinnogo/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6386-simmetrichnye-q-sorty-otsenivaiut-strukturnuiu-soglasovannost-s-moralnymi-v-llms/2026-06-23weekly0.7 https://korshunov.ai/en/article/6384-are-multilingual-models-actually-improving-isolating-true-cross-lingual-transfer/2026-06-23weekly0.7 https://korshunov.ai/en/article/6682-llama-cpp-release-b9767-adds-gpu-and-multi-platform-support/2026-06-23weekly0.7 https://korshunov.ai/en/article/6383-openbiorq-benchmark-for-agentic-biomedical-research-faithfulness/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6382-virgo-adaptivnoe-marshrutizatsiia-dlia-vizualnogo-poiska-i-globalnoi-pertseptsii/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6383-openbiorq-benchmark-dlia-vernosti-agentnykh-issledovanii-v-biomeditsine/2026-06-23weekly0.7 https://korshunov.ai/en/article/6382-virgo-adaptive-routing-for-visual-retrieval-and-global-perception/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6682-llama-cpp-release-b9767-dobavliaet-podderzhku-gpu-i-multiplatformennosti/2026-06-23weekly0.7 https://korshunov.ai/en/article/6380-can-llms-control-readability-in-arabic/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6380-mogut-li-llms-upravliat-chitaemostiu-na-arabskom/2026-06-23weekly0.7 https://korshunov.ai/en/article/6381-moshi-face-full-duplex-dialogue-with-facial-generation/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6381-moshi-face-poludupleksnaia-rechevaia-dialogovaia-model-s-generatsiei-lits/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6378-cfagentbench-metrika-dlia-avtonomnykh-agentov-po-stroitelstvu-i-finansam/2026-06-23weekly0.7 https://korshunov.ai/en/article/6378-cfagentbench-benchmark-for-autonomous-construction-finance-agents/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6377-otsenka-bolshikh-iazykovykh-modelei-dlia-preobrazovaniia-grafem-v-zvuki-na/2026-06-23weekly0.7 https://korshunov.ai/en/article/6379-bayesian-factorized-adaptation-for-code-switching-in-multilingual-asr/2026-06-23weekly0.7 https://korshunov.ai/en/article/6377-benchmarking-llms-for-japanese-grapheme-to-phoneme-conversion/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6379-baiesovskaia-faktorizovannaia-adaptatsiia-dlia-kod-perekliucheniia-v-modeliakh/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6374-web-data-recipe-for-medical-encoder-pretraining/2026-06-23weekly0.7 https://korshunov.ai/en/article/6375-nl2scratch-executable-benchmark-for-nl-to-scratch-generation/2026-06-23weekly0.7 https://korshunov.ai/en/article/6376-nous-a-predictive-world-model-for-long-term-agent-memory/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6376-nou-prognoziruiushchaia-model-mira-dlia-dolgosrochnoi-pamiati-agenta/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6375-nl2scratch-vypolnimyi-benchmark-dlia-generatsii-nl-to-scratch/2026-06-23weekly0.7 https://korshunov.ai/en/article/6374-web-data-recipe-for-medical-encoder-pretraining/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6372-mnozhestvennye-epistemologii-v-ii-tekhnologiiakh-iazyka/2026-06-23weekly0.7 https://korshunov.ai/en/article/6372-plural-epistemologies-in-ai-language-technology/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6371-tscognition-i-tsalign-prodvigaiut-rassuzhdenie-po-vremennym-riadam-s-llms/2026-06-23weekly0.7 https://korshunov.ai/en/article/6371-tscognition-and-tsalign-advance-time-series-reasoning-with-llms/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6373-mogut-li-modeli-rassuzhdenii-obnaruzhivat-izmeneniia-v-ikh-tsepiakh-myshleniia/2026-06-23weekly0.7 https://korshunov.ai/en/article/6373-can-reasoning-models-detect-changes-to-their-chains-of-thought/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6369-p-rag-ispolnenie-poiska-za-schet-semanticheskoi-kvantovaniia-i-transtsendentnogo/2026-06-23weekly0.7 https://korshunov.ai/en/article/6368-score-granularity-gap-in-llm-confidence-scoring/2026-06-23weekly0.7 https://korshunov.ai/en/article/6370-biomatrix-first-natively-multimodal-biological-foundation-model/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6370-biomatrix-pervyi-nativno-multimodalnyi-biologicheskii-fundamentalnyi-model/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6368-razryv-v-detalizatsii-otsenki-uverennosti-v-llm/2026-06-23weekly0.7 https://korshunov.ai/en/article/6369-p-rag-oblivious-retrieval-via-semantic-quantization-and-transcendental/2026-06-23weekly0.7 https://korshunov.ai/en/article/6367-measuring-genuine-emergent-consensus-in-llm-agent-societies/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6366-freimvork-leksicheskogo-konsensusa-pokazyvaet-chto-rasstoianie-vospriiatiia-slov/2026-06-23weekly0.7 https://korshunov.ai/en/article/6366-lexical-consensus-framework-shows-perceptual-distance-drives-word-learning/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6367-otsenka-iskrennego-voznikaiushchego-konsensusa-v-obshchestvakh-agentov-na-osnove/2026-06-23weekly0.7 https://korshunov.ai/en/article/6673-shipping-huggingface-hub-weekly-with-ai-open-tools-and-human-oversight/2026-06-23weekly0.7 https://korshunov.ai/en/article/6365-samatnext-v0-2-b-achieves-superior-curriculum-retention-in-small-code-models/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6365-samatnext-v0-2-b-dostigaet-prevoskhodnogo-sokhraneniia-kursa-v-malykh-modeliakh/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6674-u-nas-est-lokalnye-modeli-dlia-sortirovki-repozitoriia-openclaw-besplatno/2026-06-23weekly0.7 https://korshunov.ai/en/article/6674-we-got-local-models-to-triage-the-openclaw-repo-for-free/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6673-dostavka-huggingface-hub-ezhenedelno-s-ii-otkrytymi-instrumentami-i-kontrolem/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6670-omio-sozdaet-razgovornyi-turisticheskii-servis-na-osnove-ii/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6363-mixedpeft-edinaia-peft-dlia-neobuchennoi-adaptatsii-domena/2026-06-23weekly0.7 https://korshunov.ai/en/article/6363-mixedpeft-unified-peft-for-unsupervised-domain-adaptation/2026-06-23weekly0.7 https://korshunov.ai/en/article/6670-omio-builds-ai-native-conversational-travel/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6364-bolshie-iazykovye-modeli-ne-mogut-tochno-perevodit-fongbe/2026-06-23weekly0.7 https://korshunov.ai/en/article/6364-large-language-models-fail-to-translate-fongbe-accurately/2026-06-23weekly0.7 https://korshunov.ai/en/article/6362-asr-based-text-corpus-development-for-fongbe-and-hausa/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6360-babeljudge-otsenka-nadezhnosti-llm-kak-sudi-na-raznykh-iazykakh-i-v-agentov/2026-06-23weekly0.7 https://korshunov.ai/en/article/6360-babeljudge-measuring-llm-as-a-judge-reliability-across-languages-and-agent/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6361-adaptivnoe-planirovanie-dannykh-uluchshaet-obuchenie-na-osnove-reversii-dlia/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6362-razrabotka-korpusa-tekstov-na-osnove-audio-simvolnogo-raspoznavaniia-dlia-fongbe/2026-06-23weekly0.7 https://korshunov.ai/en/article/6361-adaptive-data-scheduling-improves-llm-reinforcement-learning/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6358-prilozhenie-curiosity-kak-lingvisticheskoe-vmeshatelstvo-v-obuchenie-s-llm/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6359-scitraj-tipizirovannyi-tsitatnyi-graf-s-podtverzhdeniem-utverzhdenii-dlia/2026-06-23weekly0.7 https://korshunov.ai/en/article/6357-orbit-training-free-multi-attribute-behavioral-steering/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6357-orbit-obuchenie-besplatnoe-upravlenie-mnogimi-atributami-povedeniia/2026-06-23weekly0.7 https://korshunov.ai/en/article/6358-curiosity-as-linguistic-intervention-in-llm-tutoring/2026-06-23weekly0.7 https://korshunov.ai/en/article/6359-scitraj-claim-grounded-typed-citation-graph-for-research-evolution/2026-06-23weekly0.7 https://korshunov.ai/en/article/6354-planbench-xl-benchmark-for-long-horizon-tool-use-planning/2026-06-23weekly0.7 https://korshunov.ai/en/article/6355-first-token-broadcasters-in-transformers-mechanistic-origins-of-language/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6354-planbench-xl-metrika-dlia-planirovaniia-ispolzovaniia-instrumentov-na-dlitelnykh/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6356-klassifikatsiia-kontseptualnoi-sinkhronizatsii-v-dialogakh-cheloveka-robota/2026-06-23weekly0.7 https://korshunov.ai/en/article/6356-a-taxonomy-of-conceptual-alignment-in-human-robot-dialogue/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6355-pervye-tokeny-v-transformerakh-mekhanicheskie-korni-identichnosti-iazyka/2026-06-23weekly0.7 https://korshunov.ai/en/article/6353-p4ir-framework-improves-llm-based-code-compliance-accuracy/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6352-zakreplenie-znanii-v-grafe-pomogaet-llm-tolko-pri-otvetakh-na-voprosy-osnovannye/2026-06-23weekly0.7 https://korshunov.ai/en/article/6352-knowledge-graph-grounding-helps-llms-only-for-out-of-training-knowledge/2026-06-23weekly0.7 https://korshunov.ai/en/article/6351-llms-use-difference-making-logic-to-learn-causal-structure/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6353-kadr-p4ir-povyshaet-tochnost-sobliudeniia-koda-na-osnove-bolshikh-iazykovykh/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6351-llms-ispolzuiut-logiku-raznosti-dlia-izucheniia-prichinnoi-struktury/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6350-raznoobrazie-personazhei-v-istoriiakh-sgenerirovannykh-llm/2026-06-23weekly0.7 https://korshunov.ai/en/article/6664-my-hugging-face-account-was-locked/2026-06-23weekly0.7 https://korshunov.ai/en/article/6350-character-variety-in-llm-generated-stories/2026-06-23weekly0.7 https://korshunov.ai/en/article/6349-speech-text-models-latently-transcribe-speech-in-intermediate-layers/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6349-modeli-rechi-teksta-latentno-transkribiruiut-rech-v-promezhutochnykh-sloiakh/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6664-moi-akkaunt-hugging-face-byl-zablokirovan/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6663-sistema-buddy-monitor-nestabilnosti-na-iazyke-rust-s-kontrolem-neopredelennosti/2026-06-23weekly0.7 https://korshunov.ai/en/article/6663-buddy-system-rust-entropy-monitor-with-ner-gated-uncertainty-for-tiered-llm/2026-06-23weekly0.7 https://korshunov.ai/en/article/6348-factor-enables-adaptive-verification-for-factuality-in-long-form-generation/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6348-factor-obespechivaet-adaptivnuiu-proverku-faktichnosti-v-generatsii-dlinnykh/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6345-variantno-kalibrovannaia-moduliatsiia-dlia-dekodirovaniia-llm/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6346-vadaorchestra-neirosimvolicheskaia-orkestratsiia-adaptivnykh-rabochikh-potokov/2026-06-23weekly0.7 https://korshunov.ai/en/article/6346-vadaorchestra-neurosymbolic-orchestration-of-adaptive-reasoning-workflows/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6347-romeva-rasshirenie-slovaria-s-sokhraneniem-geometrii-dlia-modelei-iazyka-urdu/2026-06-23weekly0.7 https://korshunov.ai/en/article/6347-romeva-geometry-preserving-vocabulary-expansion-for-roman-urdu-language-models/2026-06-23weekly0.7 https://korshunov.ai/en/article/6345-variance-calibrated-modulation-for-llm-decoding/2026-06-23weekly0.7 https://korshunov.ai/en/article/6344-gazer-training-free-semantic-correction-for-autoregressive-visual-models/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6343-macagentbench-zapuskaet-benchmark-ai-agenta-dlia-macos/2026-06-23weekly0.7 https://korshunov.ai/en/article/6343-macagentbench-launches-macos-ai-agent-benchmark/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6342-multimodalnyi-tsepochka-myslei-vozmozhnosti-i-ogranicheniia/2026-06-23weekly0.7 https://korshunov.ai/en/article/6342-multimodal-chain-of-thought-capabilities-and-limitations/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6344-gazer-obuchenie-bezopasnaia-korrektsiia-semantiki-dlia-avtoburovykh-vizualnykh/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6341-kontseptualno-ogranichennoe-obuchenie-promptov-dlia-adaptatsii-clip-pri-primerov/2026-06-23weekly0.7 https://korshunov.ai/en/article/6340-key-factors-in-rl-for-llm-reasoning-revealed/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6339-kontekstualno-osoznannoe-distilliatsiia-i-ablatsiia-dlia-text2dsl/2026-06-23weekly0.7 https://korshunov.ai/en/article/6341-concept-constrained-prompt-learning-for-few-shot-clip-adaptation/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6340-kliuchevye-faktory-v-rl-dlia-logicheskogo-myshleniia-v-llm-raskryty/2026-06-23weekly0.7 https://korshunov.ai/en/article/6339-context-aware-distillation-and-ablation-for-text2dsl/2026-06-23weekly0.7 https://korshunov.ai/en/article/6338-small-language-models-outperform-frontier-llms-in-relation-extraction/2026-06-23weekly0.7 https://korshunov.ai/en/article/6337-large-scale-cuneiform-sign-detection-dataset-and-ocr-pipeline/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6337-obemnyi-nabor-dannykh-po-raspoznavaniiu-znakov-kundrif-i-paiplain-ocr/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6235-servery-mcp-raskhoduiut-okno-konteksta-cherez-opredeleniia-instrumentov/2026-06-23weekly0.7 https://korshunov.ai/en/article/6235-mcp-servers-consume-context-window-via-tool-definitions/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6338-malye-iazykovye-modeli-prevoskhodiat-peredovye-llm-v-izvlechenii-sviazei/2026-06-23weekly0.7 https://korshunov.ai/en/article/6232-topic-to-timestamp-alignment-by-constrained-evidence-selection/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6232-soglasovanie-temy-s-vremennoi-metkoi-za-schet-vybora-dokazatelstv/2026-06-23weekly0.7 https://korshunov.ai/en/article/6231-peercheck-improving-llm-generated-academic-reviews/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6230-nalog-na-tokeny-epistemicheskoi-tochnosti-v-iskusstvennom-intellekte-osnovannom/2026-06-23weekly0.7 https://korshunov.ai/en/article/6230-the-token-tax-of-epistemic-accuracy-in-document-grounded-ai/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6229-derevia-siuzhetov-ierarkhicheskie-predstavleniia-dlia-dlinnykh-povestvovanii/2026-06-23weekly0.7 https://korshunov.ai/en/article/6229-storyline-trees-hierarchical-representations-for-long-form-narratives/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6231-peercheck-uluchshenie-akademicheskikh-otzyvov-generiruemykh-llm/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6227-ispolzovanie-vnutrennikh-artefaktov-llm-dlia-povysheniia-nadezhnosti-v-pravovoi/2026-06-23weekly0.7 https://korshunov.ai/en/article/6655-improve-user-message-detection-and-create-checkpoints/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6655-uluchshenie-obnaruzheniia-soobshchenii-polzovatelei-i-sozdanie-kontrolnykh/2026-06-23weekly0.7 https://korshunov.ai/en/article/6228-latent-personal-memory-dynamic-soft-prompts-for-llm-personalization/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6228-skrytaia-lichnaia-pamiat-dinamicheskie-miagkie-prompty-dlia-personalizatsii-llm/2026-06-23weekly0.7 https://korshunov.ai/en/article/6227-using-llm-internal-artifacts-to-improve-legal-classification-reliability/2026-06-23weekly0.7 https://korshunov.ai/en/article/6226-token-level-comparison-of-transformers-and-hybrid-models/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6226-sravnenie-na-urovne-tokenov-transformerov-i-gibridnykh-modelei/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6225-vigil3d-obespechivaet-generatsiiu-raznoobraznykh-iazykovykh-zaprosov-dlia-3d/2026-06-23weekly0.7 https://korshunov.ai/en/article/6225-vigil3d-enables-diverse-language-generation-for-3d-visual-grounding/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6224-lre-agentskaia-pamiat-na-neskolko-kilobait-s-nulevymi-neironnymi-zatratami/2026-06-23weekly0.7 https://korshunov.ai/en/article/6224-lre-few-kilobytes-agent-memory-with-zero-neural-cost/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6223-metod-napravleniia-na-etape-testirovaniia-ustraniaet-konflikty-vremennykh-faktov/2026-06-23weekly0.7 https://korshunov.ai/en/article/6223-test-time-steering-resolves-temporal-fact-conflicts-in-llms/2026-06-23weekly0.7 https://korshunov.ai/en/article/6222-ipa-based-tokenization-improves-multilingual-language-model-performance/2026-06-23weekly0.7 https://korshunov.ai/en/article/6221-beaver-agent-harness-for-scientific-curation-from-multimodal-sources/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6221-beaver-agent-interfeis-dlia-nauchnoi-kuratorskoi-raboty-iz-multimodalnykh/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6222-ispolzovanie-mezhdunarodnogo-foneticheskogo-alfavita-dlia-tokenizatsii-modelei/2026-06-23weekly0.7 https://korshunov.ai/en/article/6220-metanym-game-self-contained-llm-benchmark-for-structural-intelligence/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6220-igra-metanym-avtonomnyi-benchmark-dlia-strukturnogo-intellekta/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6219-llm-sdaiutsia-obmanu-chashche-chem-liudi/2026-06-23weekly0.7 https://korshunov.ai/en/article/6219-llms-fall-for-deception-more-than-humans/2026-06-23weekly0.7 https://korshunov.ai/en/article/6218-concepte-llm-enhanced-event-ontology-expansion/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6217-demograficheskie-metadannye-negativno-vliiaiut-na-otsenku-esse-na-osnove/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6218-concepte-rasshirenie-sobytiinoi-semantiki-s-ispolzovaniem-bolshikh-iazykovykh/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6216-mnogoetiketkovaia-annotatsiia-emotsii-analiz-soglasovannosti-i-miagkikh-golosov/2026-06-23weekly0.7 https://korshunov.ai/en/article/6217-demographic-metadata-harms-distilbert-essay-scoring/2026-06-23weekly0.7 https://korshunov.ai/en/article/6216-multilabel-emotion-annotation-agreement-and-soft-voting-analysis/2026-06-23weekly0.7 https://korshunov.ai/en/article/6215-film-coordinated-dual-branch-transformer-for-language-modeling/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6215-film-koordinirovannyi-dvoinoi-vetvevoi-transformator-dlia-modelirovaniia-iazyka/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6214-otter-sistema-krasnogo-testirovaniia-dlia-optimizatsii-promptov-izbegaiushchikh/2026-06-23weekly0.7 https://korshunov.ai/en/article/6214-otter-red-teaming-system-for-toxicity-evading-jailbreak-prompt-optimization/2026-06-23weekly0.7 https://korshunov.ai/en/article/6211-grag-framework-decouples-grounding-and-personalization-in-conversational-ai/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6211-freimvork-grag-razdeliaet-obosnovanie-i-personalizatsiiu-v-razgovornykh-ii/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6213-validatsionno-ogranichennyi-mekhanizm-analiza-vyiavleniia-suitsidalnykh-v-llms/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6212-kharakternye-vnimatelnye-transformery-dlia-obnaruzheniia-mnogokhodovykh/2026-06-23weekly0.7 https://korshunov.ai/en/article/6213-validation-gated-mechanistic-analysis-of-suicidality-detection-in-llms/2026-06-23weekly0.7 https://korshunov.ai/en/article/6212-hierarchical-attention-transformers-for-multi-turn-jailbreak-detection/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6202-oshibochnye-illiuzii-v-gollandskikh-llms/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6209-otvetnaia-inzheneriia-lokalnoe-redaktirovanie-traektorii-dlia-priniatiia-s/2026-06-23weekly0.7 https://korshunov.ai/en/article/6210-llm-based-multi-reference-evaluation-for-phrase-break-annotations/2026-06-23weekly0.7 https://korshunov.ai/en/article/6209-answer-engineering-local-trajectory-editing-for-protocol-constrained-decision/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6210-otsenka-na-osnove-llm-dlia-annotatsii-razbieniia-fraz/2026-06-23weekly0.7 https://korshunov.ai/en/article/6202-coherence-illusions-in-dutch-llms-revealed/2026-06-23weekly0.7 https://korshunov.ai/en/article/6207-adamem-learning-what-to-remember-for-personalized-long-horizon-llm-agents/2026-06-23weekly0.7 https://korshunov.ai/en/article/6208-multi-agent-audit-framework-for-clinical-mental-health-screening/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6208-mnogoagentnaia-sistema-audita-dlia-klinicheskogo-psikhicheskogo-zdorovia/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6207-adamem-obuchenie-togo-chto-nuzhno-zapominat-dlia-personalizirovannykh-agentov-na/2026-06-23weekly0.7 https://korshunov.ai/en/article/6206-study-finds-ai-still-fails-to-detect-legal-citation-hallucinations/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6206-issledovanie-pokazalo-chto-ii-vse-eshche-nesetsia-obnaruzhivat-khaltsinatsii-v/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6205-dementia-agents-multimodalnyi-multimodalnyi-sistemnyi-podkhod-k-stadirovaniiu/2026-06-23weekly0.7 https://korshunov.ai/en/article/6205-dementia-agents-multi-modal-multi-agent-system-for-dementia-staging/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6204-medlayxplain-otsenka-razryva-mezhdu-ekspertnymi-i-neprofessionalnymi-znaniiami-v/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6203-profilnyi-referent-v-fundamentalizatsii-llm/2026-06-23weekly0.7 https://korshunov.ai/en/article/6204-medlayxplain-benchmarking-expert-lay-gap-in-medical-vision-language-models/2026-06-23weekly0.7 https://korshunov.ai/en/article/6203-profile-based-reference-in-llm-grounding/2026-06-23weekly0.7 https://korshunov.ai/en/article/6199-scope-sequential-conformal-probing-for-ood-rejection-in-llms/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6201-openwer-uluchshenie-otsenki-mezhiazykovoi-rechevoi-raspoznavatelnoi-sistemy/2026-06-23weekly0.7 https://korshunov.ai/en/article/6201-openwer-enhancing-cross-lingual-asr-evaluation/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6199-scope-posledovatelnoe-konformalnoe-issledovanie-dlia-otkloneniia-vkhodov-za-v/2026-06-23weekly0.7 https://korshunov.ai/en/article/6200-rope-does-not-prevent-retrieval-heads-study-finds/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6200-rope-ne-predotvrashchaet-izvlechennye-golovki-issledovanie-pokazalo/2026-06-23weekly0.7 https://korshunov.ai/en/article/6197-listenable-interpretable-speaker-embeddings/2026-06-23weekly0.7 https://korshunov.ai/en/article/6196-synthetic-audio-framework-improves-atc-speech-recognition/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6196-sinteticheskaia-audiofreimvork-uluchshaet-raspoznavanie-rechi-v-sistemakh/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6198-arco-adaptivnyi-rubrikator-s-kogerentnym-razvitiem-dlia-agentov-na-osnove-llm/2026-06-23weekly0.7 https://korshunov.ai/en/article/6198-arco-adaptive-rubric-with-co-evolution-for-multi-step-llm-agents/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6197-slushaemye-interpretiruemye-vektory-rechevykh-embeddingov/2026-06-23weekly0.7 https://korshunov.ai/en/article/6194-factual-retrieval-in-llms-is-non-contiguous-and-redundant/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6195-otsenochnaia-ramka-dlia-vosstanovleniia-golosa-v-sistemakh-tekst-v-golos/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6194-fakticheskoe-izvlechenie-v-llms-iavliaetsia-nepreryvnym-i-izbytochnym/2026-06-23weekly0.7 https://korshunov.ai/en/article/6195-evaluation-framework-for-tts-voice-reconstruction/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6193-nauchnaia-tonkaia-nastroika-uvelichivaet-fantasticheskie-suzhdeniia-iazykovykh/2026-06-23weekly0.7 https://korshunov.ai/en/article/6192-sexualised-ai-voices-amplify-gender-power-asymmetries/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6192-seksualizirovannye-golosa-ii-usilivaiut-gendernye-nesbalansirovannye-otnosheniia/2026-06-23weekly0.7 https://korshunov.ai/en/article/6193-scientific-fine-tuning-increases-llm-hallucinations/2026-06-23weekly0.7 https://korshunov.ai/en/article/6191-cat-translate-compact-japanese-english-models-outperform-multilingual-ones-in/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6191-cat-translate-kompaktnye-modeli-iaponskogo-angliiskogo-perevoda-prevoskhodiat-v/2026-06-23weekly0.7 https://korshunov.ai/en/article/6189-benchmark-evaluation-of-small-language-models-for-arabic-nlp/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6189-otsenka-benchmarka-malykh-iazykovykh-modelei-dlia-arabskoi-nlp/2026-06-23weekly0.7 https://korshunov.ai/en/article/6190-precision-recall-controllable-radiology-report-generation/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6190-kontroliruemoe-generirovanie-meditsinskikh-otchetov-s-pomoshchiu-tochnosti-i/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6188-ekonomicheskie-sdvigi-i-kulturnaia-evoliutsiia-v-frantsuzskoi-drame/2026-06-23weekly0.7 https://korshunov.ai/en/article/6188-economic-shifts-and-cultural-evolution-in-french-drama/2026-06-23weekly0.7 https://korshunov.ai/en/article/6186-medhal-loc-benchmark-tests-localization-faithfulness-in-medical-hallucination/2026-06-23weekly0.7 https://korshunov.ai/en/article/6187-two-stage-alignment-improves-math-tutoring-pedagogy/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6187-dvukhetapnaia-sinkhronizatsiia-uluchshaet-pedagogiku-matematicheskogo-repetitora/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6186-medkhal-lok-test-model-dlia-proverki-tochnosti-lokalizatsii-oshibok-v/2026-06-23weekly0.7 https://korshunov.ai/en/article/6185-ablation-study-of-agentic-rag-components-with-local-7b-model/2026-06-23weekly0.7 https://korshunov.ai/en/article/6184-peermathdial-first-dataset-on-student-student-math-problem-solving/2026-06-23weekly0.7 https://korshunov.ai/en/article/6183-case-specific-dynamic-rubric-framework-for-translation-evaluation/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6184-peermathdial-pervyi-dataset-po-resheniiu-matematicheskikh-zadach-v-parakh-mezhdu/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6185-issledovanie-ustraneniia-komponentov-agentnogo-rag-s-ispolzovaniem-lokalnoi-7b/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6183-dinamicheskaia-sistema-otsenki-na-osnove-konkretnykh-sluchaev-dlia-otsenki/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6181-xai-zapuskaet-goal-v-grok-build-dlia-avtonomnogo-programmirovaniia/2026-06-23weekly0.7 https://korshunov.ai/en/article/6181-xai-launches-goal-in-grok-build-for-autonomous-coding/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6180-model-muzyki-na-osnove-ii-rabotaet-v-realnom-vremeni-na-bolshinstve-protsessorov/2026-06-23weekly0.7 https://korshunov.ai/en/article/6180-ai-music-model-runs-in-real-time-on-most-cpus-in-browser/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6177-vnedrenie-prompta-kak-putanitsa-rolei/2026-06-23weekly0.7 https://korshunov.ai/en/article/6177-prompt-injection-as-role-confusion/2026-06-23weekly0.7 https://korshunov.ai/en/article/6179-microsoft-releases-open-source-fastcontext-for-llm-coding-agents/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6179-microsoft-vypustil-otkrytyi-iskhodnyi-kod-fastcontext-dlia-agentov-kodirovaniia/2026-06-23weekly0.7 https://korshunov.ai/ru/article/6175-perenos-modeli-moebius-0-2b-dlia-inpainting-izobrazhenii-v-brauzer-s-claude-code/2026-06-23weekly0.7 https://korshunov.ai/en/article/6175-porting-moebius-0-2b-image-inpainting-to-browser-with-claude-code/2026-06-23weekly0.7 https://korshunov.ai/en/article/6174-gemma-4-s-potential-to-outperform-mistral-and-qwen3-6-through-finetuning/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6174-potentsial-gemma-4-prevzoiti-mistral-i-qwen3-6-za-schet-tonkoi-nastroiki/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6171-obiasnenie-riskov-ai-atak-i-vvoda-promptov/2026-06-22weekly0.7 https://korshunov.ai/en/article/6171-ai-red-teaming-and-prompt-injection-risks-explained/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6170-llama-cpp-release-b9763-dobavliaet-id-v-otvety-na-vyzovy-instrumentov/2026-06-22weekly0.7 https://korshunov.ai/en/article/6170-llama-cpp-release-b9763-adds-id-to-tool-call-responses/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6165-novyi-operator-abliatsii-kontrastnyi-kovektornyi-redaktor/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6164-deepseek-privlekaet-74-mlrd-dollarov-pri-otsenke-v-60-mlrd-dollarov-lian-venfen/2026-06-22weekly0.7 https://korshunov.ai/en/article/6164-deepseek-raises-7-4b-at-60b-valuation-liang-wenfeng-invests-3b/2026-06-22weekly0.7 https://korshunov.ai/en/article/6165-new-ablation-operator-contrastive-co-vector-edit/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6156-ideia-zapuska-glm2-pri-priemlemoi-kvantovoi-nastroike-s-gpu-i-ddr3/2026-06-22weekly0.7 https://korshunov.ai/en/article/6156-idea-for-running-glm2-at-decent-quant-with-gpu-and-ddr3-setup/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6157-primechaniia-po-vypusku-claude-v2-1-186/2026-06-22weekly0.7 https://korshunov.ai/en/article/6157-claude-v2-1-186-release-notes/2026-06-22weekly0.7 https://korshunov.ai/en/article/6146-qt-creator-20-introduces-local-ai-features/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6147-macbook-pro-m5-pro-64gb-dlia-lokalnoi-inferentsii-ii/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6146-qt-creator-20-vvodit-funktsii-lokalnoi-ii/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6145-sakana-ai-predstavliaet-sakana-fugu-model-orkestratsii-agentov/2026-06-22weekly0.7 https://korshunov.ai/en/article/6147-macbook-pro-m5-pro-64gb-for-local-ai-inference/2026-06-22weekly0.7 https://korshunov.ai/en/article/6145-sakana-ai-launches-sakana-fugu-multi-agent-orchestration-model/2026-06-22weekly0.7 https://korshunov.ai/en/article/6143-deepmind-announces-exodus-of-researchers-and-new-orchestration-models/2026-06-22weekly0.7 https://korshunov.ai/en/article/6144-glm-5-2-is-the-new-best-open-model/2026-06-22weekly0.7 https://korshunov.ai/en/article/6142-hugging-face-spaces-proxy-strips-cors-credentials-header-on-options-preflight/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6144-glm-5-2-novyi-luchshii-otkrytyi-model/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6143-deepmind-obiavliaet-ukhod-kliuchevykh-issledovatelei-i-novye-modeli-orkestratsii/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6142-proksi-hugging-face-spaces-udaliaet-zagolovok-cors-credentials-v-zaprosakh/2026-06-22weekly0.7 https://korshunov.ai/en/article/6139-add-laguna-m-1-gguf-support-by-empty-quiver/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6141-ne-udaetsia-zagruzit-veb-komponenty-gradio/2026-06-22weekly0.7 https://korshunov.ai/en/article/6141-gradio-web-components-failing-to-load/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6140-benchhub-vypustil-krupnoe-obnovlenie-dlia-otkrytogo-prostranstva-liderskikh/2026-06-22weekly0.7 https://korshunov.ai/en/article/6140-benchhub-ships-major-update-to-open-leaderboard-space/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6139-dobavlenie-podderzhki-laguna-m-1-gguf-cherez-empty-quiver/2026-06-22weekly0.7 https://korshunov.ai/en/article/6138-glm-5-2-vs-claude-opus-comparison/2026-06-22weekly0.7 https://korshunov.ai/en/article/6137-anthropic-pov-on-localllama/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6137-pozitsiia-anthropic-po-localllama/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6138-glm-5-2-protiv-claude-opus/2026-06-22weekly0.7 https://korshunov.ai/en/article/6135-why-no-thinker-models-use-sentence-level-tokens/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6135-pochemu-ni-odin-modelnyi-razrabotchik-ne-ispolzuet-tokeny-na-urovne-predlozhenii/2026-06-22weekly0.7 https://korshunov.ai/en/article/6134-jason-liu-uses-codex-for-long-running-project-management/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6134-jason-liu-ispolzuet-codex-dlia-upravleniia-dlitelnymi-proektami/2026-06-22weekly0.7 https://korshunov.ai/en/article/6129-top-n-sigma-remove-unconditional-softmax-sort/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6129-top-n-sigma-udalit-neuslovnyi-softmax-sort/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6119-openai-predstavil-instrumenty-bezopasnosti-daybreak/2026-06-22weekly0.7 https://korshunov.ai/en/article/6119-openai-launches-daybreak-security-tools/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6120-llama-cpp-release-b9761-zagruzka-modeli-i-obnovleniia-platformy/2026-06-22weekly0.7 https://korshunov.ai/en/article/6120-llama-cpp-release-b9761-model-download-and-platform-updates/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6118-openai-zapustil-patch-the-planet-dlia-otkrytogo-iskhodnogo-koda/2026-06-22weekly0.7 https://korshunov.ai/en/article/6118-openai-launches-patch-the-planet-for-open-source/2026-06-22weekly0.7 https://korshunov.ai/en/article/6112-best-local-model-for-converting-text-to-structured-json-output/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6112-luchshaia-lokalnaia-model-dlia-preobrazovaniia-teksta-v-strukturirovannyi-json/2026-06-22weekly0.7 https://korshunov.ai/en/article/6111-llama-cpp-release-b9760-new-input-schema-and-cross-platform-binaries/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6111-llama-cpp-release-b9760-novaia-skhema-vkhodnykh-dannykh-i-binarniki-dlia/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6110-cccl-runtime-sovremennyi-c-runtime-dlia-cuda/2026-06-22weekly0.7 https://korshunov.ai/en/article/6110-cccl-runtime-a-modern-c-runtime-for-cuda/2026-06-22weekly0.7 https://korshunov.ai/en/article/6108-chinese-hackers-create-tesla-v100-v4-clone/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6108-kitaiskie-khakery-sozdali-klon-tesla-v100-v4/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6109-moebius-0-2b-lightweight-image-inpainting-framework/2026-06-22weekly0.7 https://korshunov.ai/en/article/6109-moebius-0-2b-lightweight-image-inpainting-framework/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6103-nex-n2-mini-utverzhdaet-dostizhenie-pareto-optimalnosti-v-effektivnosti/2026-06-22weekly0.7 https://korshunov.ai/en/article/6104-tmax-a-simple-recipe-for-terminal-agents/2026-06-22weekly0.7 https://korshunov.ai/en/article/6103-nex-n2-mini-claims-pareto-optimality-in-reasoning-efficiency/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6104-tmax-prostoi-retsept-dlia-agentov-v-terminale/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6099-odinakovaia-model-odinakovyi-prompt-4-raznykh-agenta-daiut-raznuiu-kachestvo/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6100-gemma4-12b-qat-uncensored-balanced-released-with-60-speed-boost-via-mtp/2026-06-22weekly0.7 https://korshunov.ai/en/article/6100-gemma4-12b-qat-uncensored-balanced-released-with-60-speed-boost-via-mtp/2026-06-22weekly0.7 https://korshunov.ai/en/article/6099-same-model-same-prompt-4-different-agents-produce-varied-code-quality/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6094-glm-5-2-shag-vpered-dlia-otkrytykh-agentov/2026-06-22weekly0.7 https://korshunov.ai/en/article/6094-glm-5-2-is-the-step-change-for-open-agents/2026-06-22weekly0.7 https://korshunov.ai/en/article/6092-glm-5-2-ud-iq1-m-speed-test-on-llama-cpp-with-5090-and-3090-ti/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6092-proverka-skorosti-glm-5-2-ud-iq1-m-na-llama-cpp-s-5090-i-3090-ti/2026-06-22weekly0.7 https://korshunov.ai/en/article/6093-enable-real-time-ai-for-high-speed-data-acquisition-with-daqiri/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6093-vkliuchit-realnoe-vremia-ii-dlia-vysokoskorostnogo-sbora-dannykh-s-pomoshchiu/2026-06-22weekly0.7 https://korshunov.ai/en/article/6089-which-chassis-are-ya-using/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6089-kakoi-karkas-vy-ispolzuete/2026-06-22weekly0.7 https://korshunov.ai/en/article/6082-ai-out-persuades-humans-new-study-shows-ai-superior-to-experts/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6084-nvidia-predstavliaet-halos-dlia-robototekhniki-polnosistemnyi-funktsionalnyi/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6082-ii-prevoskhodit-liudei-novoe-issledovanie-pokazyvaet-chto-ii-prevoskhodit/2026-06-22weekly0.7 https://korshunov.ai/en/article/6084-nvidia-launches-halos-for-robotics-full-stack-functional-safety-system/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6081-ia-sozdal-instrument-kotoryi-pozvoliaet-izbegat-ruchnogo-perekliucheniia-modelei/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6083-pp-ocrv6-vypushchen-na-hugging-face-s-podderzhkoi-50-iazykov/2026-06-22weekly0.7 https://korshunov.ai/en/article/6083-pp-ocrv6-released-on-hugging-face-with-50-language-support/2026-06-22weekly0.7 https://korshunov.ai/en/article/6081-i-built-a-tool-to-stop-manually-swapping-models-on-my-8gb-gpu/2026-06-22weekly0.7 https://korshunov.ai/en/article/6080-glm5-2-runs-at-7tg-on-4x-3090s-with-192gb-ddr5-on-budget-build/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6080-glm5-2-rabotaet-na-7tg-na-4kh-geforce-3090-s-192-gb-ddr5-v-biudzhetnom-sborke/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6073-qwen3-6-35b-a3b-apex-na-rtx-3090-sravnitelnye-pokazateli-skorosti-i-kachestva/2026-06-22weekly0.7 https://korshunov.ai/en/article/6073-qwen3-6-35b-a3b-apex-on-rtx-3090-speed-and-quality-benchmarks/2026-06-22weekly0.7 https://korshunov.ai/en/article/6072-pci-passthrough-only-hits-gen-1-speed/2026-06-22weekly0.7 https://korshunov.ai/en/article/6068-seeking-indic-document-datasets-for-ai-ocr-training-in-india/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6068-poisk-naborov-dokumentov-dlia-obucheniia-ii-ocr-na-indiiskikh-iazykakh-v-indii/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6072-pci-passthrough-tolko-dostigaet-skorosti-pokoleniia-1/2026-06-22weekly0.7 https://korshunov.ai/en/article/6071-eu-ddr5-prices-dropping-germany-much-cheaper-than-netherlands/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6071-tseny-na-ddr5-v-es-snizhaiutsia-v-germanii-oni-znachitelno-deshevle-chem-v/2026-06-22weekly0.7 https://korshunov.ai/en/article/6069-moonmath-ai-open-sources-hip-attention-kernel-that-beats-aiter-v3-on-mi300x/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6069-moonmath-ai-opublikoval-otkrytyi-iskhodnyi-kod-iadra-pristavki-vnimaniia-hip-v3/2026-06-22weekly0.7 https://korshunov.ai/en/article/6066-the-starting-status-bug-is-now-fully-fixed/2026-06-22weekly0.7 https://korshunov.ai/en/article/6067-space-flagged-as-abusive-by-cloudflare-rule/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6067-prostranstvo-pomecheno-kak-vredonosnoe-po-pravilu-cloudflare/2026-06-22weekly0.7 https://korshunov.ai/en/article/6063-i-built-a-novel-triple-hybrid-llm-under-1b-parameters-for-50/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6063-ia-sozdal-novuiu-troinuiu-gibridnuiu-model-iazykovoi-modeli-s-menee-chem-1-za-50/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6066-bag-s-statusom-starting-teper-polnostiu-ustranen/2026-06-22weekly0.7 https://korshunov.ai/en/article/6064-what-ai-trend-will-have-the-biggest-impact-in-the-next-five-years/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6065-ml-modeli-zameshcheniia-v-cfd-fea-prakticheskie-podkhody-i-problemy/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6064-kakoi-trend-v-ii-okazhet-naibolshee-vliianie-v-blizhaishie-piat-let/2026-06-22weekly0.7 https://korshunov.ai/en/article/6065-ml-surrogate-models-in-cfd-fea-real-world-practices-and-challenges/2026-06-22weekly0.7 https://korshunov.ai/en/article/6062-llama-cpp-release-b9757-new-binaries-and-features/2026-06-22weekly0.7 https://korshunov.ai/en/article/6061-llama-cpp-release-b9758-adds-bf16-support-and-broader-compiler-compatibility/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6062-llama-cpp-release-b9757-new-binaries-and-features/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6061-llama-cpp-release-b9758-dobavliaet-podderzhku-bf16-i-rasshirennuiu-sovmestimost/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6060-net-evropeiskikh-postavshchikov-inferensa-dlia-glm-5-2-ili-deepseek-v4-flash/2026-06-22weekly0.7 https://korshunov.ai/en/article/6060-no-european-inference-providers-for-glm-5-2-or-deepseek-v4-flash/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6059-qat-kv-cache-quantization-dlia-gemma-4-31b-pokazyvaet-ogromnoe-uluchshenie/2026-06-22weekly0.7 https://korshunov.ai/en/article/6059-qat-kv-cache-quantization-for-gemma-4-31b-shows-massive-improvement/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6058-otchet-po-tekhnicheskim-kharakteristikam-ling-i-ring-2-6-vypusk-modelei-s-v/2026-06-22weekly0.7 https://korshunov.ai/en/article/6058-ling-and-ring-2-6-technical-report-releases-trillion-parameter-models/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6053-mozhet-li-jetson-orin-nano-zapuskat-kodovyi-model-takoi-kak-qwen/2026-06-22weekly0.7 https://korshunov.ai/en/article/6053-can-jetson-orin-nano-run-a-coding-model-like-qwen/2026-06-22weekly0.7 https://korshunov.ai/en/article/6052-gemma-4-qat-31b-responds-better-to-kv-cache-quantization/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6052-gemma-4-qat-31b-luchshe-reagiruet-na-kvantovanie-kesha-kv/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6050-ispravlenie-sboia-pri-redaktirovanii-faila-pri-dobavlenii-v-konets-faila/2026-06-22weekly0.7 https://korshunov.ai/en/article/6050-fix-for-edit-file-crash-on-append-at-file-end/2026-06-22weekly0.7 https://korshunov.ai/en/article/6047-support-for-step3-5-3-7-flash-mtp3-added/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6047-dobavleno-podderzhka-flash-mtp3-step3-5-3-7/2026-06-22weekly0.7 https://korshunov.ai/en/article/6046-gemma-4-31b-q6-runs-at-8-9-t-s-on-dual-9060-xt-cards/2026-06-22weekly0.7 https://korshunov.ai/en/article/6044-will-dedicated-hardware-for-local-llms-become-affordable-soon/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6044-skoro-li-stanet-dostupnym-spetsializirovannoe-oborudovanie-dlia-lokalnykh/2026-06-22weekly0.7 https://korshunov.ai/ru/article/6046-gemma-4-31b-q6-rabotaet-so-skorostiu-8-9-tokenov-v-sekundu-na-dvukh-kartakh-9060/2026-06-22weekly0.7 https://korshunov.ai/en/article/5940-tutorial-on-building-python-first-interactive-dashboards-with-prefab/2026-06-22weekly0.7 https://korshunov.ai/en/article/5941-the-7-types-of-agent-memory-a-technical-guide/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5940-instruktsiia-po-sozdaniiu-interaktivnykh-dashbordov-na-python-s-ispolzovaniem/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5941-7-tipov-pamiati-agenta-tekhnicheskoe-rukovodstvo/2026-06-22weekly0.7 https://korshunov.ai/en/article/5937-infinitely-stuck-on-starting-with-docker-container-running/2026-06-22weekly0.7 https://korshunov.ai/en/article/5938-the-clockwork-dark-a-local-first-ai-narrative-rpg-engine/2026-06-22weekly0.7 https://korshunov.ai/en/article/5939-capability-is-not-in-the-weights-empirical-negative-result-on-mlp-weight/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5939-navyk-ne-zakliuchen-v-vesakh-empiricheskii-otritsatelnyi-rezultat-po-proektsii/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5938-tma-iz-mekhanizmov-lokalnyi-pervyi-dvizhok-ai-rolevoi-igry-s-siuzhetom/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5937-beskonechno-zastrevaet-na-zapuske-s-konteinerom-docker-rabotaiushchim-s-r-shiny/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5934-llm-kak-epistemicheskie-uskoriteli-risk-zakliuchaetsia-ne-tolko-v/2026-06-22weekly0.7 https://korshunov.ai/en/article/5935-small-scale-debug-comparison-of-olmo-core-with-engram-graft/2026-06-22weekly0.7 https://korshunov.ai/en/article/5934-llms-as-epistemic-accelerators-the-risk-is-not-only-hallucination/2026-06-22weekly0.7 https://korshunov.ai/en/article/5936-nova-vad-beats-silero-pyannote-and-webrtc-on-noisy-audio-with-93-accuracy/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5936-nova-vad-pobezhdaet-silero-pyannote-i-webrtc-pri-shumnom-audio-s-tochnostiu-93/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5935-malomasshtabnoe-sravnenie-otladki-olmo-core-s-gratom-engram/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5931-shablony-dlia-sozdaniia-otsenok-v-oblasti-kiberbezopasnosti/2026-06-22weekly0.7 https://korshunov.ai/en/article/5932-space-stuck-restarting-on-old-commit-for-16-hours/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5932-prostranstvo-zastrialo-na-perezapuske-v-staroi-versii-kommita-na-16-chasov/2026-06-22weekly0.7 https://korshunov.ai/en/article/5931-patterns-for-building-cybersecurity-evals/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5933-desiatistorentnye-kartochki-dlia-uskoreniia-ii-dostupny/2026-06-22weekly0.7 https://korshunov.ai/en/article/5933-tenstorrent-ai-accelerator-cards-available/2026-06-22weekly0.7 https://korshunov.ai/en/article/5911-llama-cpp-release-b9753-new-binaries-and-progress-reporting/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5913-llama-cpp-release-b9751-new-binaries-and-fixes/2026-06-22weekly0.7 https://korshunov.ai/en/article/5913-llama-cpp-release-b9751-new-binaries-and-fixes/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5912-llama-cpp-release-b9752-server-refactor-and-cross-platform-binaries/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5911-llama-cpp-release-b9753-new-binaries-and-progress-reporting/2026-06-22weekly0.7 https://korshunov.ai/en/article/5912-llama-cpp-release-b9752-server-refactor-and-cross-platform-binaries/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5903-sravnenie-modelei-gemma-4-31b-q6-i-gemma-4-31b-qat/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5910-llama-cpp-release-b9754-novyi-parser-ac-i-binarniki-dlia-raznykh-platform/2026-06-22weekly0.7 https://korshunov.ai/en/article/5910-llama-cpp-release-b9754-new-ac-parser-and-cross-platform-binaries/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5909-dobavit-zavisimost-libandroid-spawn-dlia-sborki-android/2026-06-22weekly0.7 https://korshunov.ai/en/article/5909-add-libandroid-spawn-dependency-for-android-build/2026-06-22weekly0.7 https://korshunov.ai/en/article/5903-gemma-4-31b-q6-vs-gemma-4-31b-qat-comparison/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5901-rabotaiushchii-protsess-dlia-programmistov-s-medlennoi-lokalnoi-nastroikoi-llm/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5902-sravnenie-lokalnykh-modelei-tekst-k-izobrazheniiu-okonchatelnoe-ispytanie/2026-06-22weekly0.7 https://korshunov.ai/en/article/5902-local-text-to-image-model-comparison-the-ultimate-test/2026-06-22weekly0.7 https://korshunov.ai/en/article/5900-your-favorite-workflow-to-convert-pdf-with-complex-structure-to-markdown/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5900-vash-liubimyi-rabochii-protsess-preobrazovaniia-pdf-s-slozhnoi-strukturoi-v/2026-06-22weekly0.7 https://korshunov.ai/en/article/5901-workflow-for-programmers-with-slow-local-llm-setup/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5899-rekomendatsii-po-nastroike-python-proekta-veb-prilozheniia/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5898-nakonets-to-vidim-vygody-mtp-posle-udaleniia-ggml-cuda-allreduce/2026-06-22weekly0.7 https://korshunov.ai/en/article/5898-finally-seeing-benefits-of-mtp-after-removing-ggml-cuda-allreduce/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5897-agent-hermes-vygliadit-uzhasno-i-imeet-plokhoe-vzaimodeistvie-s-polzovatelem/2026-06-22weekly0.7 https://korshunov.ai/en/article/5897-hermes-agent-looks-ugly-and-has-poor-ux/2026-06-22weekly0.7 https://korshunov.ai/en/article/5899-agent-recommendations-for-python-web-project-setup/2026-06-22weekly0.7 https://korshunov.ai/en/article/5894-local-llm-inference-optimization-the-complete-guide/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5894-optimizatsiia-inferensa-lokalnykh-modelei-iazykovoi-modeli-polnoe-rukovodstvo/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5896-tablitsa-rezultatov-dlia-kvantovannykh-modelei-podobnykh-analizu-iskusstvennogo/2026-06-22weekly0.7 https://korshunov.ai/en/article/5896-leaderboard-for-quantized-models-similar-to-artificial-analysis/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5895-ne-novyi-model-prosto-prazdnik-ottsa-i-blagodarnost/2026-06-22weekly0.7 https://korshunov.ai/en/article/5895-not-a-new-model-just-a-happy-father-s-day-and-a-thank-you/2026-06-22weekly0.7 https://korshunov.ai/en/article/5893-glm-5-2-released-on-deepswe-benchmark/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5892-samsung-vnedriaet-chatgpt-i-codex-dlia-sotrudnikov/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5893-vypushchena-versiia-glm-5-2-na-benchmarke-deepswe/2026-06-22weekly0.7 https://korshunov.ai/en/article/5892-samsung-deploys-chatgpt-and-codex-for-employees/2026-06-22weekly0.7 https://korshunov.ai/en/article/5870-cloudflare-launches-temporary-accounts-for-ai-agents/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5870-cloudflare-zapuskaet-vremennye-akkaunty-dlia-ii-agentov/2026-06-22weekly0.7 https://korshunov.ai/en/article/5869-sqlite-utils-4-0rc1-release/2026-06-22weekly0.7 https://korshunov.ai/en/article/5868-sqlite-utils-4-0rc1-adds-migrations-and-nested-transactions/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5868-sqlite-utils-4-0rc1-dobavliaet-migratsii-i-vlozhennye-tranzaktsii/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5869-reliz-sqlite-utils-4-0rc1/2026-06-22weekly0.7 https://korshunov.ai/ru/article/5838-qwen-27b-dlia-planirovaniia-qwen-35b-a3b-dlia-vypolneniia/2026-06-21weekly0.7 https://korshunov.ai/en/article/5838-qwen-27b-for-planning-qwen-35b-a3b-for-execution/2026-06-21weekly0.7 https://korshunov.ai/en/article/5836-llama-cpp-release-b9750-new-call-statement-and-cross-platform-binaries/2026-06-21weekly0.7 https://korshunov.ai/en/article/5835-updated-vision-model-benchmark-results-and-recommendations/2026-06-21weekly0.7 https://korshunov.ai/ru/article/5836-llama-cpp-release-b9750-novaia-instruktsiia-vyzova-i-binarniki-dlia-raznykh/2026-06-21weekly0.7 https://korshunov.ai/en/article/5834-qwen-3-6-27b-apostate-released-with-safety-removed/2026-06-21weekly0.7 https://korshunov.ai/ru/article/5835-rezultaty-i-rekomendatsii-po-obnovlennomu-benchmarku-vizualnykh-modelei/2026-06-21weekly0.7 https://korshunov.ai/ru/article/5834-vypushchena-versiia-qwen-3-6-27b-s-udaleniem-bezopasnosti/2026-06-21weekly0.7 https://korshunov.ai/en/article/5831-i-forked-ik-llama-cpp-and-added-numa-mirror-mode/2026-06-21weekly0.7 https://korshunov.ai/ru/article/5831-ia-forknul-ik-llama-cpp-i-dobavil-rezhim-numa-mirror/2026-06-21weekly0.7 https://korshunov.ai/ru/article/5825-llama-cpp-vypuskaet-versiiu-b9748-s-novymi-binarnikami-i-funktsiiami/2026-06-21weekly0.7 https://korshunov.ai/en/article/5825-llama-cpp-releases-version-b9748-with-new-binaries-and-features/2026-06-21weekly0.7 https://korshunov.ai/en/article/5824-i-pretrained-and-post-trained-a-500m-parameter-llm-and-330m-parameter-image-from/2026-06-21weekly0.7 https://korshunov.ai/ru/article/5824-ia-predobuchil-i-postobuchil-500m-parametrovuiu-model-llm-i-330m-parametrovuiu-s/2026-06-21weekly0.7 https://korshunov.ai/ru/article/5820-kakoi-vash-lokalnyi-analog-haiku/2026-06-21weekly0.7 https://korshunov.ai/en/article/5820-what-s-your-local-haiku-replacement/2026-06-21weekly0.7 https://korshunov.ai/ru/article/5777-gpt-5-6-artefakty-claude-code-predstavlennaia-pamiat-brain-ot-perplexity/2026-06-21weekly0.7 https://korshunov.ai/en/article/5777-gpt-5-6-claude-code-artifacts-perplexity-s-brain-memory-released/2026-06-21weekly0.7 https://korshunov.ai/ru/article/5496-zapros-na-prinuditelnoe-udalenie-zastriavshego-prostranstva-hugging-face/2026-06-21weekly0.7 https://korshunov.ai/en/article/4574-glm-5-2-breakout-and-open-model-progress-highlighted/2026-06-21weekly0.7 https://korshunov.ai/ru/article/4574-razryv-i-progress-otkrytiia-modelei-glm-5-2/2026-06-21weekly0.7 https://korshunov.ai/en/article/5496-request-to-force-delete-stuck-hugging-face-space/2026-06-21weekly0.7 https://korshunov.ai/en/article/3571-build-your-own-transaction-foundation-model-for-financial-intelligence/2026-06-21weekly0.7 https://korshunov.ai/ru/article/3571-sozdaite-sobstvennyi-fundamentalnuiu-model-tranzaktsii-dlia-finansovoi/2026-06-21weekly0.7 https://korshunov.ai/en/article/3771-can-you-beat-lora-in-fine-tuning/2026-06-21weekly0.7 https://korshunov.ai/ru/article/3771-mozhno-li-pobedit-lora-pri-melkom-nastroike/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2408-gpt-5-5-instant-uluchshaet-otvety-chatgpt-po-voprosam-zdorovia/2026-06-21weekly0.7 https://korshunov.ai/ru/article/3420-plan-upravleniia-ii-dlia-bezopasnosti-vnutrennikh-sistem/2026-06-21weekly0.7 https://korshunov.ai/en/article/3420-ai-control-roadmap-for-internal-system-security/2026-06-21weekly0.7 https://korshunov.ai/en/article/2408-gpt-5-5-instant-enhances-chatgpt-s-health-responses/2026-06-21weekly0.7 https://korshunov.ai/en/article/5495-important-finding-for-everyone-stuck-on-starting-status/2026-06-21weekly0.7 https://korshunov.ai/en/article/5459-new-research-shows-amie-matches-doctors-in-disease-management/2026-06-21weekly0.7 https://korshunov.ai/ru/article/5495-vazhnoe-otkrytie-dlia-vsekh-kto-zastrial-na-statuse-zapusk/2026-06-21weekly0.7 https://korshunov.ai/ru/article/4573-glm-5-2-poiavliaetsia-kak-vedushchaia-model-otkrytogo-vesa-dlia-kodirovaniia/2026-06-21weekly0.7 https://korshunov.ai/ru/article/5459-novoe-issledovanie-pokazyvaet-chto-amie-sootvetstvuet-vracham-po-upravleniiu/2026-06-21weekly0.7 https://korshunov.ai/en/article/4573-glm-5-2-emerges-as-leading-open-weight-coding-model/2026-06-21weekly0.7 https://korshunov.ai/ru/article/3419-pravitelstvo-velikobritanii-i-google-deepmind-zapustili-prototip-planirovaniia/2026-06-21weekly0.7 https://korshunov.ai/en/article/3770-mosaicleaks-can-your-research-agent-keep-a-secret/2026-06-21weekly0.7 https://korshunov.ai/en/article/3570-nvidia-launches-xr-ai-for-ar-glasses-and-wearable-devices/2026-06-21weekly0.7 https://korshunov.ai/ru/article/3770-mosaicleaks-mozhet-li-vash-issledovatelskii-agent-sokhraniat-sekret/2026-06-21weekly0.7 https://korshunov.ai/en/article/3419-uk-government-and-google-deepmind-launch-ai-powered-housing-planning-prototype/2026-06-21weekly0.7 https://korshunov.ai/ru/article/3570-nvidia-predstavliaet-xr-ai-dlia-ar-okularov-i-nosimykh-ustroistv/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2407-openai-zapuskaet-kontrol-raskhodov-i-analitiku-ispolzovaniia-dlia-chatgpt/2026-06-21weekly0.7 https://korshunov.ai/en/article/2407-openai-launches-spend-controls-and-usage-analytics-for-chatgpt-enterprise/2026-06-21weekly0.7 https://korshunov.ai/ru/article/5525-iskliuchitelno-skidka-250-dollarov-na-bilety-na-konferentsiiu-ai-engineer-do/2026-06-21weekly0.7 https://korshunov.ai/en/article/5525-exclusive-250-off-ai-engineer-tickets-until-monday/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2406-2x-radeon-r9700-s-modeliu-qwen-3-6-27b-q8-mtp-na-llama-cpp/2026-06-21weekly0.7 https://korshunov.ai/en/article/2406-2x-radeon-r9700-with-qwen-3-6-27b-q8-mtp-on-llama-cpp/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2394-tokenicheskaia-model-na-reddit-localllama/2026-06-21weekly0.7 https://korshunov.ai/en/article/2394-tokenomics-post-on-localllama-reddit/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2391-mogu-li-ia-realno-dostich-sposobnostei-klaude-kodeksa-lokalno/2026-06-21weekly0.7 https://korshunov.ai/en/article/2391-can-i-realistically-get-close-to-claude-codex-capabilities-locally/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2390-rocm-protiv-vulkan-protiv-vllm-po-proizvoditelnosti-na-dvoinykh-r9700/2026-06-21weekly0.7 https://korshunov.ai/en/article/2390-rocm-vs-vulkan-vs-vllm-performance-on-dual-r9700s/2026-06-21weekly0.7 https://korshunov.ai/en/article/2388-llama-cpp-release-b9747-adds-real-time-model-load-tracking-and-new-platform/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2388-llama-cpp-release-b9747-dobavliaet-otslezhivanie-zagruzki-modeli-v-realnom-i/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2387-sandboksirovanie-vypolneniia-koda-dlia-agentov-ii/2026-06-21weekly0.7 https://korshunov.ai/en/article/2387-sandboxing-code-execution-for-ai-agents/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2381-llama-cpp-release-b9745-dobavliaet-podderzhku-mtp3-i-binarniki-dlia-raznykh/2026-06-21weekly0.7 https://korshunov.ai/en/article/2381-llama-cpp-release-b9745-adds-mtp3-support-and-cross-platform-binaries/2026-06-21weekly0.7 https://korshunov.ai/en/article/2380-running-mimo-2-5-on-two-halo-strixeses/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2380-zapusk-mimo-2-5-na-dvukh-mashinakh-halo-strixese/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2371-klod-skoro-potrebuet-podtverzhdeniia-lichnosti/2026-06-21weekly0.7 https://korshunov.ai/en/article/2374-thinking-loop-bug-in-opencode-with-local-model/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2374-oshibki-tsikla-myshleniia-v-opencode-s-lokalnymi-modeliami/2026-06-21weekly0.7 https://korshunov.ai/en/article/2375-8-16-mi50s-minimax-m3-19-tps-tg-peak/2026-06-21weekly0.7 https://korshunov.ai/en/article/2371-claude-will-soon-require-identity-verification/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2375-8-16-mi50s-minimax-m3-19-tps-tg-peak/2026-06-21weekly0.7 https://korshunov.ai/en/article/2372-r9700-gpu-performance-issues-with-vllm-and-multi-gpu-setup/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2372-problemy-s-proizvoditelnostiu-gpu-r9700-pri-ispolzovanii-vllm-i-nastroike-gpu/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2366-pochemu-autoround-tak-silno-ignoriruetsia/2026-06-21weekly0.7 https://korshunov.ai/en/article/2366-why-is-autoround-being-slept-on-so-hard/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2369-ia-otslezhival-kazhduiu-konfiguratsionnuiu-fail-dlia-agenta-i-pometil-realnoe/2026-06-21weekly0.7 https://korshunov.ai/en/article/2369-i-mapped-every-agent-config-file-and-tagged-real-adoption/2026-06-21weekly0.7 https://korshunov.ai/en/article/2367-proposal-for-splitting-base-models-to-avoid-retraining/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2367-predlozhenie-o-razdelenii-bazovykh-modelei-dlia-izbezhaniia-pereobucheniia/2026-06-21weekly0.7 https://korshunov.ai/en/article/2364-glm-5-2-beats-gemini-and-gpt-5-4-in-coding-but-is-inefficient/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2364-glm-5-2-pobezhdaet-gemini-i-gpt-5-4-v-programmirovanii-no-iavliaetsia/2026-06-21weekly0.7 https://korshunov.ai/en/article/2365-watch-local-llms-escape-the-rooms-you-design/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2365-sledite-za-tem-kak-lokalnye-llm-vykhodiat-iz-komnat-kotorye-vy-sozdaete/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2362-gemma-4-qat-luchshe-reagiruet-na-kvantovanie-kesha-kv/2026-06-21weekly0.7 https://korshunov.ai/en/article/2362-gemma-4-qat-responds-better-to-kv-cache-quantization/2026-06-21weekly0.7 https://korshunov.ai/en/article/2358-fable-vs-glm-5-2-vs-kimi-k2-7-youtube-video/2026-06-21weekly0.7 https://korshunov.ai/en/article/2357-vercel-ceo-says-almost-shocked-by-glm-5-2-s-coding-abilities/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2358-fable-protiv-glm-5-2-protiv-kimi-k2-7-youtube-video/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2357-vitse-prezident-vercel-govorit-chto-pochti-shokirovano-ot-sposobnostei-glm-5-2-v/2026-06-21weekly0.7 https://korshunov.ai/en/article/2355-qwen-3-7-will-not-be-open-sourced/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2355-qwen-3-7-ne-budet-otkrytoi-versii/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2348-predlagaemaia-model-chuvstv-ispolzuet-tolko-emodzi/2026-06-21weekly0.7 https://korshunov.ai/en/article/2348-proposed-feeling-model-uses-only-emojis/2026-06-21weekly0.7 https://korshunov.ai/en/article/2346-kimi-ai-just-mailed-me/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2346-kimi-ai-prosto-otpravil-mne-pismo/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2345-allenai-vypustil-modeli-molmomotion-dlia-prognozirovaniia-dvizhenii-v-budushchem/2026-06-21weekly0.7 https://korshunov.ai/en/article/2345-allenai-releases-molmomotion-vision-models-for-future-motion-prediction/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2341-supralabs-zapuskaet-semeistvo-modelei-any2any/2026-06-21weekly0.7 https://korshunov.ai/en/article/2341-supralabs-launches-any2any-model-family/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2340-chto-vy-pereproektiruete-na-chto-nikto-nikogda-ne-budet-ispolzovat-budte-chestny/2026-06-21weekly0.7 https://korshunov.ai/en/article/2340-what-are-you-overengineering-that-nobody-s-ever-going-to-use-be-honest/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2339-reliz-llama-cpp-b9744-novye-binarniki-i-funktsii/2026-06-21weekly0.7 https://korshunov.ai/en/article/2339-llama-cpp-release-b9744-new-binaries-and-features/2026-06-21weekly0.7 https://korshunov.ai/en/article/2337-best-open-source-vision-model-runnable-on-rtx-6000-pro/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2337-luchshaia-otkrytaia-model-vizualnogo-analiza-rabotaiushchaia-na-rtx-6000-pro/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2334-semantic-memory-lokalnaia-baza-znanii-s-tipizirovannymi-grafovymi-rebrami/2026-06-21weekly0.7 https://korshunov.ai/en/article/2334-semantic-memory-local-first-knowledge-base-with-typed-graph-edges/2026-06-21weekly0.7 https://korshunov.ai/en/article/2332-what-can-i-run-on-my-tesla-v100-32gb-system/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2332-chto-ia-mogu-zapustit-na-moei-sisteme-tesla-v100-32gb/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2331-chto-delaiut-liudi-s-lokalnymi-modeliami-i-kakie-instrumenty-oni-ispolzuiut/2026-06-21weekly0.7 https://korshunov.ai/en/article/2331-what-are-people-doing-with-their-local-models-and-what-tools-do-they-use/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2327-chto-proiskhodit-kogda-podpiski-na-llm-perestaiut-podderzhivatsia/2026-06-21weekly0.7 https://korshunov.ai/en/article/2327-what-happens-when-llm-subscriptions-stop-being-subsidized/2026-06-21weekly0.7 https://korshunov.ai/en/article/2324-anyone-running-minimax-m3-with-pipenetwork-mixed-3-6-quant/2026-06-21weekly0.7 https://korshunov.ai/ru/article/2324-kto-nibud-zapuskaet-minimax-m3-s-pipenetwork-mixed-3-6-quant/2026-06-21weekly0.7 https://korshunov.ai/en/article/2323-introducing-noema-atlas-peer-to-peer-model-distribution/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2323-vvedenie-v-noema-atlas-raspredelenie-modelei-na-osnove-pir-top-pir/2026-06-20weekly0.7 https://korshunov.ai/en/article/2316-llama-cpp-release-b9743-adds-new-binaries-and-gpu-support/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2317-llama-cpp-release-b9742-obnovleno-s-ispolzovaniem-zapolnennogo-shaga-dlia-vesov/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2316-llama-cpp-release-b9743-adds-new-binaries-and-gpu-support/2026-06-20weekly0.7 https://korshunov.ai/en/article/2317-llama-cpp-release-b9742-updated-with-padded-stride-for-ssm-conv-weights/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2300-obnovleno-podskazka-o-prostoiakh-potoka-v-v2-1-185/2026-06-20weekly0.7 https://korshunov.ai/en/article/2300-stream-stall-hint-updated-in-v2-1-185/2026-06-20weekly0.7 https://korshunov.ai/en/article/2298-glm-5-2-local-inference-speeds-report/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2298-glom-5-2-otchet-o-skorosti-lokalnoi-inferentsii/2026-06-20weekly0.7 https://korshunov.ai/en/article/2299-six-months-ago-i-turned-down-8165-for-an-rtx-6000-pro/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2299-mesiats-nazad-ia-otkazalsia-ot-8165-dollarov-za-rtx-6000-pro/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2290-llama-cpp-release-b9741-adds-new-binaries-and-support/2026-06-20weekly0.7 https://korshunov.ai/en/article/2290-llama-cpp-release-b9741-adds-new-binaries-and-support/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2289-besplatnaia-seriia-iz-15-chastei-o-vnutrennikh-aspektakh-llm-osnovannykh-na-4/2026-06-20weekly0.7 https://korshunov.ai/en/article/2289-free-15-part-series-on-llm-internals-grounded-in-gemma-4-12b/2026-06-20weekly0.7 https://korshunov.ai/en/article/2288-qwen-code-companion-extension-now-open-sourced/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2288-rasshirenie-qwen-code-companion-teper-otkryto-dlia-ispolzovaniia/2026-06-20weekly0.7 https://korshunov.ai/en/article/2287-gemma-4-26b-a4b-excels-in-language-and-scientific-queries/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2287-gemma-4-26b-a4b-vydeliaetsia-pri-otvetakh-na-iazykovye-i-nauchnye-zaprosy/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2285-slozhnosti-s-zaversheniem-token-planov-xiaomi-mimo-v2-5-pro-do-istecheniia-sroka/2026-06-20weekly0.7 https://korshunov.ai/en/article/2285-struggling-to-finish-xiaomi-mimo-v2-5-pro-token-plan-credits-before-expiry/2026-06-20weekly0.7 https://korshunov.ai/en/article/2281-board-where-every-tile-is-an-agent/2026-06-20weekly0.7 https://korshunov.ai/en/article/2282-fix-for-test-args-parser-random-failures-on-windows/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2281-doska-gde-kazhdyi-kvadrat-agent/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2282-ispravlenie-sluchainykh-sboev-v-test-args-parser-na-windows/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2273-glubokaia-neironnaia-set-prevrashchaet-izobrazheniia-v-igrovye-sessii-lokalno/2026-06-20weekly0.7 https://korshunov.ai/en/article/2273-deep-neural-network-turns-images-into-playable-games-locally/2026-06-20weekly0.7 https://korshunov.ai/en/article/2272-r9700-purchase-decision-amid-gpu-price-surge/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2272-reshenie-o-pokupke-r9700-v-usloviiakh-rosta-tsen-na-videokarty/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2269-sovety-2x-3090/2026-06-20weekly0.7 https://korshunov.ai/en/article/2269-advice-2x-3090/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2265-teper-vy-mozhete-konvertirovat-kvantovannye-modeli-exl3-na-mac-s-protsessorom/2026-06-20weekly0.7 https://korshunov.ai/en/article/2265-you-can-now-convert-exl3-quants-on-apple-silicon-mac/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2259-llama-cpp-release-b9739-adds-win-opencl-adreno-arm64-support/2026-06-20weekly0.7 https://korshunov.ai/en/article/2259-llama-cpp-release-b9739-adds-win-opencl-adreno-arm64-support/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2247-dva-dokumenta-word-obmenivaiutsia-soderzhimym-cherez-lokalnye-llms-realnye/2026-06-20weekly0.7 https://korshunov.ai/en/article/2248-napkin-math-on-collective-hosting-costs-for-diffusiongemma-in-2026/2026-06-20weekly0.7 https://korshunov.ai/en/article/2247-two-word-docs-chatting-via-local-llms-real-use-cases/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2248-napisannoe-na-listke-bumagi-matematika-po-sovokupnym-raskhodam-na-khosting-v/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2242-llama-cpp-release-b9738-fixes-cors-auth-header-forwarding-and-new-binary-builds/2026-06-20weekly0.7 https://korshunov.ai/en/article/2242-llama-cpp-release-b9738-fixes-cors-auth-header-forwarding-and-new-binary-builds/2026-06-20weekly0.7 https://korshunov.ai/en/article/2241-any-opinion-about-qwen3-6-27b-bf16-vs-step3-7-iq4-xs/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2241-kakoi-iz-modelei-luchshe-qwen3-6-27b-bf16-ili-step3-7-iq4-xs/2026-06-20weekly0.7 https://korshunov.ai/en/article/2240-z-ai-praises-the-number-1-open-source-model/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2240-z-ai-khvalit-pervyi-otkrytyi-model/2026-06-20weekly0.7 https://korshunov.ai/en/article/2237-research-project-injecting-natural-language-tactical-intent-into-multi-agent/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2237-issledovatelskii-proekt-vnedrenie-estestvennogo-iazykovogo-strategicheskogo-v/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2233-glm-5-2-ud-iq2-m-sozdaet-luchshii-izobrazhenie-pelikana-v-formate-svg-kotoroe/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2234-luchshii-lokalnyi-llm-dlia-summirovaniia-angliiskikh-rasskazov/2026-06-20weekly0.7 https://korshunov.ai/en/article/2234-best-local-llm-for-english-story-summarization/2026-06-20weekly0.7 https://korshunov.ai/en/article/2233-glm-5-2-ud-iq2-m-produces-best-pelican-svg-image-ever-seen/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2228-ggml-optimiziruet-amx-s-pomoshchiu-ploskoi-obrabotki-razdelov/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2227-ispravlenie-indeksatora-dsa-dlia-glm-5-2-tenzory-pomecheny-kak-ne-trebuemye/2026-06-20weekly0.7 https://korshunov.ai/en/article/2227-glm-5-2-dsa-indexer-fix-tensors-marked-not-required/2026-06-20weekly0.7 https://korshunov.ai/en/article/2228-ggml-optimizes-amx-with-partition-flattening/2026-06-20weekly0.7 https://korshunov.ai/en/article/2226-docker-prebuild-web-ui-for-s390x/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2226-docker-predvaritelnoe-postroenie-veb-interfeisa-dlia-s390x/2026-06-20weekly0.7 https://korshunov.ai/en/article/2223-worlds-biggest-chat-title-dataset-released-by-supralabs/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2223-vypushchena-samaia-bolshaia-v-mire-baza-dannykh-zagolovkov-chatov-ot-supralabs/2026-06-20weekly0.7 https://korshunov.ai/en/article/2220-latent-space-subscribers-get-250-discount-for-aie-wf-2026/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2220-poluchateli-prostranstva-skrytykh-poluchaiut-skidku-v-razmere-250-dollarov-za-wf/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2219-nailuchshie-nastroiki-dlia-48-gb-ozu-s-qwen-3-6-27b/2026-06-20weekly0.7 https://korshunov.ai/en/article/2219-best-settings-for-48gb-vram-with-qwen-3-6-27b/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2215-glm-5-2-dostigaet-98-maksimalnoi-intellekta-s-menee-chem-polovinoi-tokenov/2026-06-20weekly0.7 https://korshunov.ai/en/article/2216-7900xtx-24gb-vram-runs-qwen-3-6-27b-with-131k-context/2026-06-20weekly0.7 https://korshunov.ai/en/article/2217-help-with-a-local-document-rag-system-storage-ingestion-query-highlighting/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2217-pomoshch-s-lokalnoi-sistemoi-rag-dlia-dokumentov-khranilishche-vvod-zapros/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2216-7900xtx-24gb-vram-runs-qwen-3-6-27b-s-131k-context/2026-06-20weekly0.7 https://korshunov.ai/en/article/2215-glm-5-2-achieves-98-max-intelligence-with-less-than-half-tokens/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2207-amd-budushchie-gpu-dlia-razvertyvaniia-llm/2026-06-20weekly0.7 https://korshunov.ai/en/article/2207-amd-future-gpu-offerings-for-llm-builds/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2206-rezultaty-otsenok-llama-cpp-b70-s-ispolzovaniem-sycl/2026-06-20weekly0.7 https://korshunov.ai/en/article/2206-llama-cpp-b70-sycl-benchmarks-results/2026-06-20weekly0.7 https://korshunov.ai/en/article/2204-local-ai-for-local-office-files/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2204-lokalnaia-ii-dlia-lokalnykh-failov-ofisa/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2202-problema-vyzova-instrumentov-v-otkrytom-qwen3-6-27b-8k/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2201-kakaia-luchshaia-kniga-dlia-izucheniia-matematiki-v-oblasti-mashinnogo-i/2026-06-20weekly0.7 https://korshunov.ai/en/article/2202-tool-calling-issue-in-open-source-qwen3-6-27b-8k/2026-06-20weekly0.7 https://korshunov.ai/en/article/2201-what-is-the-best-book-for-learning-ml-deep-learning-maths/2026-06-20weekly0.7 https://korshunov.ai/en/article/2199-rust-release-0-0-15/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2199-rust-0-0-15/2026-06-20weekly0.7 https://korshunov.ai/en/article/2198-open-interpreter-0-0-16-released/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2195-lokalnyi-agent-dostupa-k-veb-saitam-cherez-searxng-i-scrapling/2026-06-20weekly0.7 https://korshunov.ai/en/article/2195-local-agent-web-access-via-searxng-and-scrapling/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2197-opublikovana-versiia-open-interpreter-0-0-17/2026-06-20weekly0.7 https://korshunov.ai/en/article/2197-open-interpreter-0-0-17-released/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2198-opublikovano-open-interpreter-0-0-16/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2196-lokalnyi-agent-na-4090-poisk-nastroek-lm-studio/2026-06-20weekly0.7 https://korshunov.ai/en/article/2196-local-agent-on-4090-looking-for-lm-studio-settings/2026-06-20weekly0.7 https://korshunov.ai/en/article/2192-rtx-5090-msi-power-usage-and-cable-warning/2026-06-20weekly0.7 https://korshunov.ai/en/article/2193-supralabs-releases-supra-title-fft-preview-with-115k-samples/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2193-supralabs-vypustil-supra-title-fft-preview-s-115k-obraztsov/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2192-rtx-5090-msi-potreblenie-energii-i-preduprezhdenie-o-kabele/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2187-vnimanie-algebra-grammatika-perevodiashchaia-estestvennyi-iazyk-v-spektrogrammy/2026-06-20weekly0.7 https://korshunov.ai/en/article/2187-attention-algebra-a-grammar-that-translates-natural-language-into-spectrograms/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2181-reliz-llama-cpp-b9732-novye-binarniki-i-obnovleniia/2026-06-20weekly0.7 https://korshunov.ai/en/article/2181-llama-cpp-release-b9732-new-binaries-and-updates/2026-06-20weekly0.7 https://korshunov.ai/en/article/2183-i-benchmarked-claude-s-fast-c-it-wasn-t-faster/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2183-ia-protestiroval-bystryi-c-klaude-on-ne-byl-bystree/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2180-ggml-webgpu-dobavliaet-perekliuchateli-adaptera-dlia-podderzhki-f16-na-vulkan-i/2026-06-20weekly0.7 https://korshunov.ai/en/article/2180-ggml-webgpu-adds-f16-adapter-toggles-for-vulkan-and-nvidia/2026-06-20weekly0.7 https://korshunov.ai/en/article/2175-1800-gpu-cost-runs-qwen3-6-27b-with-262k-context-and-55-tok-s/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2175-stoimost-v-1800-gpu-pozvoliaet-zapustit-qwen3-6-27b-s-kontekstom-v-262k-i-55-tok/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2170-sean-linch-o-izoliatsii-potokov-autentifikatsii-v-mcp/2026-06-20weekly0.7 https://korshunov.ai/en/article/2170-sean-lynch-on-mcp-s-auth-flow-isolation/2026-06-20weekly0.7 https://korshunov.ai/en/article/2165-llama-cpp-release-b9731-performance-optimization-and-cross-platform-binaries/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2165-llama-cpp-release-b9731-optimizatsiia-proizvoditelnosti-i-binarniki-dlia/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2166-llama-cpp-release-b9730-fixes-and-new-binaries/2026-06-20weekly0.7 https://korshunov.ai/en/article/2166-llama-cpp-release-b9730-fixes-and-new-binaries/2026-06-20weekly0.7 https://korshunov.ai/en/article/2161-best-local-agents-jun-2026/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2161-luchshie-lokalnye-agenty-iiun-2026/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2157-rust-0-0-13/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2158-rust-0-0-12/2026-06-20weekly0.7 https://korshunov.ai/en/article/2158-rust-release-0-0-12/2026-06-20weekly0.7 https://korshunov.ai/en/article/2157-rust-release-0-0-13/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2156-rust-0-0-14/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2155-pomoshch-v-zapuske-lokalnogo-agenta-hermes-s-ispolzovaniem-llama-cpp/2026-06-20weekly0.7 https://korshunov.ai/en/article/2155-help-running-local-hermes-agent-with-llama-cpp/2026-06-20weekly0.7 https://korshunov.ai/en/article/2156-rust-release-0-0-14/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2149-maksimizatsiia-proizvoditelnosti-dvukh-3090-s-ispolzovaniem-nvlink/2026-06-20weekly0.7 https://korshunov.ai/en/article/2149-maximizing-performance-of-2x3090-with-nvlink/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2150-reliz-llama-cpp-b9729-novye-binarniki-i-podderzhka-platform/2026-06-20weekly0.7 https://korshunov.ai/en/article/2150-llama-cpp-release-b9729-new-binaries-and-platform-support/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2034-supralabs-vypustil-model-supravl-nano-900k-dlia-obrabotki-izobrazhenii-i-iazyka/2026-06-20weekly0.7 https://korshunov.ai/en/article/2034-supralabs-releases-supravl-nano-900k-vision-language-model/2026-06-20weekly0.7 https://korshunov.ai/en/article/2147-how-to-set-optimal-llama-cpp-parameters-for-amd-gpu/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2147-kak-nastroit-optimalnye-parametry-llama-cpp-dlia-modelei-gemma-4-na-gpu-ot-amd/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2143-ustranenie-problemy-s-dlitelnymi-kontekstami-pri-dekodirovanii-na-radeon-r9700-s/2026-06-20weekly0.7 https://korshunov.ai/en/article/2143-fixing-long-context-decode-cliff-on-radeon-r9700-with-vllm-0-22-1/2026-06-20weekly0.7 https://korshunov.ai/en/article/2139-how-to-setup-search-with-ai-models/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2139-kak-nastroit-poisk-s-pomoshchiu-modelei-ii/2026-06-20weekly0.7 https://korshunov.ai/ru/article/2135-reliz-llama-cpp-b9728-dobavliaet-podderzhku-strok-kommentariev-i-binarnye-faily/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2134-modeli-glm-5-2-reap50-gguf-dostupny-na-hugging-face/2026-06-19weekly0.7 https://korshunov.ai/en/article/2135-llama-cpp-release-b9728-adds-comment-line-support-and-multiple-platform-binaries/2026-06-19weekly0.7 https://korshunov.ai/en/article/2134-glm-5-2-reap50-gguf-models-available-on-hugging-face/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2128-mozhno-li-ispolzovat-ssd-dlia-rasshireniia-pamiati-bez-swap-na-mac-mini-m4/2026-06-19weekly0.7 https://korshunov.ai/en/article/2128-can-you-use-an-ssd-to-extend-memory-without-swap-on-mac-mini-m4/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2127-komissiia-vybiraet-konsortsium-europa-v-kachestve-pobeditelia-konkursa-frontier/2026-06-19weekly0.7 https://korshunov.ai/en/article/2127-commission-selects-europa-consortium-as-winner-of-frontier-ai-grande-challenge/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2124-llama-cpp-release-b9726-dobavliaet-argument-agent-i-novye-binarniki-dlia/2026-06-19weekly0.7 https://korshunov.ai/en/article/2125-improving-local-models-with-an-api-based-consultant-agent/2026-06-19weekly0.7 https://korshunov.ai/en/article/2124-llama-cpp-release-b9726-adds-agent-arg-and-new-platform-binaries/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2125-uluchshenie-lokalnykh-modelei-s-pomoshchiu-agenta-konsultanta-na-osnove-api/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2121-ekonomika-ii-nachinaet-otdavat-predpochtenie-otkrytym-modeliam/2026-06-19weekly0.7 https://korshunov.ai/en/article/2121-the-economics-of-ai-are-starting-to-favor-open-models/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2123-llama-cpp-release-b9727-obnovlenie-do-versii-cpp-httplib-0-48-0/2026-06-19weekly0.7 https://korshunov.ai/en/article/2123-llama-cpp-release-b9727-update-to-cpp-httplib-0-48-0/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2122-lq50-24-dostupna-angliiskaia-versiia/2026-06-19weekly0.7 https://korshunov.ai/en/article/2122-lq50-24-english-translation-available/2026-06-19weekly0.7 https://korshunov.ai/en/article/2119-claude-fable-5-and-mythos-5-capabilities/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2119-claude-fable-5-i-mythos-5-vozmozhnosti/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2118-otsenka-ili-otsenka-marketing/2026-06-19weekly0.7 https://korshunov.ai/en/article/2118-benchmarking-or-benchmarketing/2026-06-19weekly0.7 https://korshunov.ai/en/article/2117-docker-build-the-ui-24794/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2117-docker-sborka-elementa-interfeisa-24794/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2114-dobavlenie-vtorogo-gpu-na-materinskuiu-platu-x670e-dlia-lokalnykh-bolshikh/2026-06-19weekly0.7 https://korshunov.ai/en/article/2114-adding-a-second-gpu-to-x670e-motherboard-for-local-llms/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2113-luchshii-keis-dlia-poiska-v-internete/2026-06-19weekly0.7 https://korshunov.ai/en/article/2113-best-harness-for-web-searching/2026-06-19weekly0.7 https://korshunov.ai/en/article/2112-what-s-more-impressive-glm-5-1-to-5-2-or-qwen-3-5-to-3-6/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2112-chto-bolee-vpechatliaet-glm-5-1-do-5-2-ili-qwen-3-5-do-3-6/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2108-zapret-na-otkrytyi-iskhodnyi-kod-ii-budet-oshibkoi/2026-06-19weekly0.7 https://korshunov.ai/en/article/2108-banning-open-source-ai-would-be-a-mistake/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2107-ia-slishkom-slabyi-cpu-i-ram-dlia-lokalnykh-llm/2026-06-19weekly0.7 https://korshunov.ai/en/article/2107-is-my-cpu-and-ram-too-weak-for-local-llms/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2102-nabliudenie-za-tem-kak-lokalnyi-ii-assistent-po-golosu-stanovitsia-menee-umnym/2026-06-19weekly0.7 https://korshunov.ai/en/article/2103-llama-cpp-release-b9724-with-bug-fixes-and-cross-platform-binaries/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2103-reliz-llama-cpp-b9724-s-ispravleniem-oshibok-i-binarnikami-dlia-raznykh-platform/2026-06-19weekly0.7 https://korshunov.ai/en/article/2102-watching-a-local-ai-voice-assistant-get-dumber/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2101-glm-5-2-eto-novyi-vedushchii-otkrytyi-modelnyi-ves-na-indekse-iskusstvennogo/2026-06-19weekly0.7 https://korshunov.ai/en/article/2101-glm-5-2-is-the-new-leading-open-weights-model-on-the-artificial-analysis-index/2026-06-19weekly0.7 https://korshunov.ai/en/article/2098-new-agentic-benchmark-released/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2100-eagle3-poiavilsia-dlia-qwen/2026-06-19weekly0.7 https://korshunov.ai/en/article/2100-the-eagle3-has-landed-for-qwen/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2098-vypushchena-novaia-otsenka-agentov/2026-06-19weekly0.7 https://korshunov.ai/en/article/2095-spec-support-eagle3-for-qwen3-5-3-6-by-ruixiang63/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2095-spec-podderzhka-eagle3-dlia-qwen3-5-i-qwen3-6-ot-ruixiang63/2026-06-19weekly0.7 https://korshunov.ai/en/article/2096-llama-cpp-release-b9723-adds-support-for-qwen3-5-and-qwen3-6-eagle3/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2096-llama-cpp-release-b9723-dobavliaet-podderzhku-qwen3-5-i-qwen3-6-eagle3/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2093-kto-to-ispolzoval-vibethinker-3b-vne-benchmarkov/2026-06-19weekly0.7 https://korshunov.ai/en/article/2093-has-anyone-used-vibethinker-3b-outside-benchmarks/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2091-kto-to-zdes-ispolzuet-dve-rtx-5090/2026-06-19weekly0.7 https://korshunov.ai/en/article/2091-anyone-here-rocking-dual-rtx-5090s/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2092-reliz-llama-cpp-b9722-ispravleniia-i-binarniki-dlia-raznykh-platform/2026-06-19weekly0.7 https://korshunov.ai/en/article/2092-llama-cpp-release-b9722-fixes-and-cross-platform-binaries/2026-06-19weekly0.7 https://korshunov.ai/en/article/2088-multi-doc-agent-workflows-in-word/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2088-mnogourovnevye-rabochie-protsessy-agentov-v-word/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2090-otmecheno-tsenzurirovanie-lokalnykh-llm-na-reddit/2026-06-19weekly0.7 https://korshunov.ai/en/article/2090-local-llm-censorship-reported-on-reddit/2026-06-19weekly0.7 https://korshunov.ai/en/article/2086-evotensile-evolutionary-tuning-of-amd-tensile-gemm-kernels/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2086-evotensile-evoliutsionnaia-nastroika-iader-amd-tensile-gemm/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2087-mem-dolzhen-prodolzhatsia/2026-06-19weekly0.7 https://korshunov.ai/en/article/2087-the-meme-must-go-on/2026-06-19weekly0.7 https://korshunov.ai/en/article/2085-ohio-state-university-releases-open-source-deep-research-agent-quest-35b/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2085-okho-seit-universitet-vypuskaet-otkrytyi-deep-research-agent-quest-35b/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2084-llama-cpp-release-b9718-consolidated-slot-selection-and-new-binary-builds/2026-06-19weekly0.7 https://korshunov.ai/en/article/2084-llama-cpp-release-b9718-consolidated-slot-selection-and-new-binary-builds/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2082-okho-seit-universitet-vypuskaet-otkrytyi-deep-research-agent-quest-35b/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2083-llama-cpp-release-b9721-dostupen-dlia-neskolkikh-platform/2026-06-19weekly0.7 https://korshunov.ai/en/article/2082-ohio-state-university-releases-open-source-deep-research-agent-quest-35b/2026-06-19weekly0.7 https://korshunov.ai/en/article/2083-llama-cpp-release-b9721-available-for-multiple-platforms/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2080-glm-5-2-teper-mozhet-rabotat-lokalno-v-llama-cpp-i-unsloth-studio/2026-06-19weekly0.7 https://korshunov.ai/en/article/2080-glm-5-2-can-now-run-locally-in-llama-cpp-and-unsloth-studio/2026-06-19weekly0.7 https://korshunov.ai/en/article/2078-ggml-cpu-adds-k-tails-support-for-power10-mma-q8-q4/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2077-malo-pomalu-spasibo-komande-deepseek/2026-06-19weekly0.7 https://korshunov.ai/en/article/2077-little-late-thank-you-to-the-deepseek-team/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2078-ggml-cpu-dobavliaet-podderzhku-k-tails-dlia-power10-mma-q8-q4/2026-06-19weekly0.7 https://korshunov.ai/en/article/2076-guys-le-chaton-fat-is-real/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2076-gospoda-le-chaton-fat-realnyi/2026-06-19weekly0.7 https://korshunov.ai/en/article/2072-v0-17-6-release-notes/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2072-primechaniia-po-vypusku-v0-17-6/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2071-glm-5-2-proshel-proverku-na-atmosferu-prevzoshel-gpt-5-5/2026-06-19weekly0.7 https://korshunov.ai/en/article/2071-glm-5-2-passes-vibe-check-outperforms-gpt-5-5/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2067-kak-ia-mogu-samostoiatelno-razvernut-instrumenty-proverki-koda/2026-06-19weekly0.7 https://korshunov.ai/en/article/2066-llama-cpp-release-b9716-adds-batching-support-for-internvl/2026-06-19weekly0.7 https://korshunov.ai/en/article/2067-how-can-i-self-host-code-review/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2066-llama-cpp-release-b9716-adds-batching-support-for-internvl/2026-06-19weekly0.7 https://korshunov.ai/en/article/2062-llama-cpp-releases-b9713-with-new-binaries-and-features/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2062-llama-cpp-vypuskaet-b9713-s-novymi-binarnikami-i-funktsiiami/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2061-llama-cpp-release-b9714-dobavliaet-zagolovok-x-accel-buffering-i-novye-binarniki/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2059-ufp4-edinaia-4-bitovaia-trenirovka-ustraniaet-smeshchenie-szhatiia-v-trenirovke/2026-06-19weekly0.7 https://korshunov.ai/en/article/2059-ufp4-uniform-4-bit-training-overcomes-shrinkage-bias-in-llm-pretraining/2026-06-19weekly0.7 https://korshunov.ai/en/article/2061-llama-cpp-release-b9714-adds-x-accel-buffering-header-and-new-binaries/2026-06-19weekly0.7 https://korshunov.ai/en/article/2060-llama-cpp-release-b9715-adds-cuda-col2im-1d-and-multiple-platform-binaries/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2060-reliz-llama-cpp-b9715-dobavliaet-podderzhku-cuda-col2im-1d-i-binarnye-faily-dlia/2026-06-19weekly0.7 https://korshunov.ai/en/article/2057-nrt-bench-multi-turn-red-teaming-of-llm-agents-in-safety-critical-systems/2026-06-19weekly0.7 https://korshunov.ai/en/article/2058-datamagic-turns-tabular-data-into-interactive-insight-videos/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2058-datamagic-prevrashchaet-tablitsy-dannykh-v-interaktivnye-video-s-analizom/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2057-nrt-bench-mnogokhodovoe-krasnoe-testirovanie-agentov-llm-v-kriticheski-vazhnykh/2026-06-19weekly0.7 https://korshunov.ai/en/article/2055-attention-guided-deep-learning-for-interpretable-sperm-morphology-classification/2026-06-19weekly0.7 https://korshunov.ai/en/article/2056-multi-view-decompilation-improves-llm-based-malware-classification/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2056-mnogourovnevaia-detsifratsiia-uluchshaet-klassifikatsiiu-vredonosnogo-po-na/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2055-vnimanie-orientirovannyi-glubokii-analiz-dlia-interpretiruemoi-klassifikatsii/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2053-kontekstualno-osoznannyi-baiesovskii-model-uluchshaet-prognozirovanie-uspekha/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2054-pereispolzovanie-klassifikatora-rechi-dlia-generatsii-na-osnove-diffuzii/2026-06-19weekly0.7 https://korshunov.ai/en/article/2053-context-aware-bayesian-model-improves-ivf-success-prediction/2026-06-19weekly0.7 https://korshunov.ai/en/article/2054-repurposing-speech-classifier-for-diffusion-based-generation/2026-06-19weekly0.7 https://korshunov.ai/en/article/2052-defensive-misdirection-against-automated-attacks-on-agentic-ai/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2051-ultraquant-4-bit-keshirovanie-kv-dlia-agentov-s-bolshim-kontekstom/2026-06-19weekly0.7 https://korshunov.ai/en/article/2051-ultraquant-4-bit-kv-caching-for-context-heavy-agents/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2052-zashchita-ot-avtomatizirovannykh-atak-na-agentnye-ii/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2049-peredacha-predvziatosti-otsenshchikov-v-sistemakh-multiagentnykh-iazykovykh/2026-06-19weekly0.7 https://korshunov.ai/en/article/2050-optimal-order-in-multi-agent-systems-framework/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2050-optimalnyi-poriadok-v-ramke-mnogoagentnykh-sistem/2026-06-19weekly0.7 https://korshunov.ai/en/article/2049-evaluator-bias-propagation-in-multi-agent-llm-systems/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2047-freestyle-masshtabiruemoe-generirovanie-dvoinykh-ssylok-cherez-izvlechenie-loras/2026-06-19weekly0.7 https://korshunov.ai/en/article/2047-freestyle-scalable-dual-reference-generation-via-community-lora-mining/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2048-kalibrovka-bez-ponimaniia-v-obnaruzhenii-uiazvimostei-v-llm/2026-06-19weekly0.7 https://korshunov.ai/en/article/2048-calibration-without-comprehension-in-llm-vulnerability-detection/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2046-kak-bezopasnye-llm-interpretiruiut-smeshannye-demonstratsii-sobliudeniia/2026-06-19weekly0.7 https://korshunov.ai/en/article/2045-efficient-and-sound-probabilistic-verification-for-ai-agents/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2045-effektivnaia-i-nadezhnaia-veroiatnostnaia-proverka-dlia-agentov-ii/2026-06-19weekly0.7 https://korshunov.ai/en/article/2046-how-safety-aligned-llms-interpret-mixed-compliance-demonstrations/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2044-multi-lcb-rasshirenie-livecodebench-do-12-iazykov-programmirovaniia/2026-06-19weekly0.7 https://korshunov.ai/en/article/2043-flowedit-lifelong-pronunciation-adaptation-in-flow-matching-tts/2026-06-19weekly0.7 https://korshunov.ai/en/article/2044-multi-lcb-extending-livecodebench-to-12-programming-languages/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2043-flowedit-dolgosrochnaia-adaptatsiia-proiznosheniia-v-modeli-flow-matching-tts/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2041-sarlo-80-vysokorazreshaiushchii-sar-opticheskii-tekstovyi-nabor-dannykh/2026-06-19weekly0.7 https://korshunov.ai/en/article/2041-sarlo-80-vhr-sar-optical-text-dataset-released/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2042-sovershenno-samostoiatelnyi-broker-ispolneniia-dlia-agentnogo-upravleniia-s/2026-06-19weekly0.7 https://korshunov.ai/en/article/2042-sovereign-execution-broker-for-certificate-bound-agentic-control/2026-06-19weekly0.7 https://korshunov.ai/en/article/2040-deepswip-counterfactual-reasoning-in-neural-probabilistic-logic/2026-06-19weekly0.7 https://korshunov.ai/en/article/2039-ledgeragent-structured-state-for-policy-adherent-tool-calling-agents/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2039-ledgeragent-strukturirovannyi-status-dlia-agentov-vyzova-instrumentov-politiku/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2040-deepswip-kontrfaktalnoe-myshlenie-v-neironnoi-veroiatnostnoi-logike/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2037-kalibrovka-v-modeliakh-na-osnove-smesi-ekspertov-pri-smeshchenii-raspredeleniia/2026-06-19weekly0.7 https://korshunov.ai/en/article/2038-cross-attention-attribution-for-style-captioned-text-to-speech/2026-06-19weekly0.7 https://korshunov.ai/en/article/2037-calibration-in-moe-models-under-distribution-shift/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2038-perekrestnoe-vnimanie-po-atributsii-dlia-stilia-opisatelnogo-teksta-k-golosu/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2036-g2rec-edinaia-platforma-dlia-generativnykh-rekomendatsii/2026-06-19weekly0.7 https://korshunov.ai/en/article/2035-how-transparent-is-diffusiongemma/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2035-na-skolko-prozrachna-diffusiongemma/2026-06-19weekly0.7 https://korshunov.ai/en/article/2036-g2rec-unified-framework-for-generative-recommendation/2026-06-19weekly0.7 https://korshunov.ai/en/article/2032-fedmgs-federated-modality-aware-graph-synthesis-for-imbalanced-multimodal/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2032-fedmgs-federativnoe-modalnost-osoznannoe-sintezirovanie-grafa-dlia-obucheniia/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2033-primechaniia-po-vypusku-v1-14-8a2/2026-06-19weekly0.7 https://korshunov.ai/en/article/2033-v1-14-8a2-release-notes/2026-06-19weekly0.7 https://korshunov.ai/en/article/2031-style-diversity-outperforms-topic-diversity-in-annotation-free-synthetic-data/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2031-raznoobrazie-stilia-prevoskhodit-raznoobrazie-tem-v-sintezirovannykh-dannykh-bez/2026-06-19weekly0.7 https://korshunov.ai/en/article/2030-direct-advantage-estimation-for-partially-observable-domains/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2028-vremennye-vlozheniia-ne-nuzhny-v-modeliakh-diffuzii/2026-06-19weekly0.7 https://korshunov.ai/en/article/2029-lightweight-defense-against-false-data-injection-in-power-grids/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2029-legkaia-zashchita-ot-vvedeniia-lozhnykh-dannykh-v-seti-elektrosnabzheniia/2026-06-19weekly0.7 https://korshunov.ai/en/article/2028-timestep-embeddings-unnecessary-in-diffusion-models/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2030-priamaia-otsenka-vygody-dlia-chastichno-nabliudaemykh-oblastei/2026-06-19weekly0.7 https://korshunov.ai/en/article/2027-deepgala-neural-surrogates-with-uncertainty-for-pde-inverse-problems/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2025-heptv2-effektivnyi-konechnyi-transformator-dlia-rekonstruktsii-zariazhennykh/2026-06-19weekly0.7 https://korshunov.ai/en/article/2025-heptv2-end-to-end-efficient-point-transformer-for-charged-particle/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2026-mekhanicheskoe-issledovanie-sokhraneniia-predstavlenii-v-nepreryvnom-obuchenii/2026-06-19weekly0.7 https://korshunov.ai/en/article/2026-mechanistic-study-of-representation-retention-in-continual-learning/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2027-deepgala-neironnye-zameniteli-s-neopredelennostiu-dlia-obratnykh-zadach-v/2026-06-19weekly0.7 https://korshunov.ai/en/article/2022-ssh-net-deep-network-for-failure-time-prediction-under-competing-risks/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2023-topologicheskii-analiz-dannykh-dlia-monitoringa-protsessov-v-realnom-vremeni/2026-06-19weekly0.7 https://korshunov.ai/en/article/2024-two-stage-evolutionary-hyperparameter-optimization-for-pinns/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2024-dvukhetapnaia-evoliutsionnaia-optimizatsiia-giperparametrov-dlia-pinns/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2022-ssh-net-glubokaia-set-dlia-prognozirovaniia-vremeni-otkaza-pri-riskakh/2026-06-19weekly0.7 https://korshunov.ai/en/article/2023-topological-data-analysis-for-real-time-process-monitoring/2026-06-19weekly0.7 https://korshunov.ai/en/article/2020-bias-mitigation-under-coverage-constraints-and-the-price-of-fairness/2026-06-19weekly0.7 https://korshunov.ai/en/article/2019-agentic-symbolic-search-for-pde-solution-characterization/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2020-ustranenie-predvziatosti-pri-ogranicheniiakh-na-pokrytie-i-tsena-spravedlivosti/2026-06-19weekly0.7 https://korshunov.ai/en/article/2021-repurposing-speech-classifier-for-diffusion-based-generation/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2021-pereispolzovanie-klassifikatora-rechi-dlia-generatsii-na-osnove-diffuzii/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2019-agentnyi-simvolicheskii-poisk-dlia-kharakteristiki-reshenii-uravnenii-v/2026-06-19weekly0.7 https://korshunov.ai/en/article/2017-ultraquant-4-bit-kv-caching-for-context-heavy-agents/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2017-ultraquant-4-bitnoe-keshirovanie-kv-dlia-agentov-s-bolshim-kontekstom/2026-06-19weekly0.7 https://korshunov.ai/en/article/2016-marginal-advantage-accumulation-for-memory-driven-agent-self-evolution/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2016-sbor-mezhgruppovogo-preimushchestva-dlia-samoevoliutsii-agentov-osnovannykh-na/2026-06-19weekly0.7 https://korshunov.ai/en/article/2018-riemannian-sharpness-explains-sgd-s-bias-toward-flat-minima/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2018-rimanovskaia-ostrota-obiasniaet-predvziatost-sgd-k-ploskim-minimumam/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2015-dataset-refrad2d-obespechivaet-masshtabiruemuiu-prostranstvennuiu-sopostavimost/2026-06-19weekly0.7 https://korshunov.ai/en/article/2013-evaluator-bias-propagation-in-multi-agent-llm-systems/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2013-peredacha-predvziatosti-v-sistemakh-multiagentnykh-iazykovykh-modelei/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2014-sovmeshchenie-llm-s-ispolzovaniem-skrytoi-obratnoi-sviazi-polzovatelia/2026-06-19weekly0.7 https://korshunov.ai/en/article/2014-llm-alignment-using-implicit-user-feedback/2026-06-19weekly0.7 https://korshunov.ai/en/article/2015-refrad2d-dataset-enables-scalable-spatial-grounding-in-radiology/2026-06-19weekly0.7 https://korshunov.ai/en/article/2010-probe-and-refine-tuning-improves-coding-agent-performance/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2011-kak-bezopasnye-llm-interpretiruiut-smeshannye-demonstratsii-sobliudeniia/2026-06-19weekly0.7 https://korshunov.ai/en/article/2011-how-safety-aligned-llms-interpret-mixed-compliance-demonstrations/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2012-otsenka-entropii-v-sistemakh-s-neskolkimi-kutritami-s-pomoshchiu-neironnykh/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2010-metod-probe-and-refine-uluchshaet-proizvoditelnost-kod-agenta/2026-06-19weekly0.7 https://korshunov.ai/en/article/2012-entropy-estimation-in-multi-qutrit-systems-with-neural-networks/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2005-prognoziruemost-kak-melkoe-izmerenie-dlia-konfidentsialnosti/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2008-kapsuly-sostoianiia-vypolneniia-dlia-nizkozaderzhannogo-vypolneniia-ii-na/2026-06-19weekly0.7 https://korshunov.ai/en/article/2008-execution-state-capsules-for-low-latency-on-device-ai-serving/2026-06-19weekly0.7 https://korshunov.ai/en/article/2005-predictability-as-a-fine-grained-measure-for-privacy/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2009-sovershenno-samostoiatelnyi-ispolnitelnyi-broker-dlia-agentnogo-kontrolia-s/2026-06-19weekly0.7 https://korshunov.ai/en/article/2009-sovereign-execution-broker-for-certificate-bound-agentic-control/2026-06-19weekly0.7 https://korshunov.ai/en/article/2007-multi-task-bayesian-in-context-learning-framework/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2006-kalibrovka-v-modeliakh-na-osnove-smesi-ekspertov-pri-smeshchenii-raspredeleniia/2026-06-19weekly0.7 https://korshunov.ai/en/article/2006-calibration-in-moe-models-under-distribution-shift/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2007-mnogotselevaia-baiesovskaia-sistema-obucheniia-v-kontekste/2026-06-19weekly0.7 https://korshunov.ai/en/article/2003-deterministic-multicalibration-with-optimal-sample-complexity/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2003-opredelennaia-mnogokalibrovka-s-optimalnoi-slozhnostiu-vyborki/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2002-uniego-proksi-sredi-edinaia-egotsentricheskaia-videopredstavlenie/2026-06-19weekly0.7 https://korshunov.ai/en/article/2004-lie-algebra-attention-group-element-tokens-in-neural-networks/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2004-li-algebraicheskoe-vnimanie-tokeny-elementov-gruppy-v-neironnykh-setiakh/2026-06-19weekly0.7 https://korshunov.ai/en/article/2002-uniego-proxy-mediated-unified-egocentric-video-representation/2026-06-19weekly0.7 https://korshunov.ai/en/article/2001-how-transparent-is-diffusiongemma/2026-06-19weekly0.7 https://korshunov.ai/en/article/2000-refrad2d-dataset-enables-scalable-spatial-grounding-in-radiology/2026-06-19weekly0.7 https://korshunov.ai/en/article/1999-llm-alignment-using-implicit-user-feedback/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1999-sovmeshchenie-llm-s-ispolzovaniem-skrytoi-obratnoi-sviazi-polzovatelia/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2001-na-skolko-prozrachna-diffusiongemma/2026-06-19weekly0.7 https://korshunov.ai/ru/article/2000-dataset-refrad2d-obespechivaet-masshtabiruemuiu-prostranstvennuiu-oporu-v/2026-06-19weekly0.7 https://korshunov.ai/en/article/1996-ledgeragent-structured-state-for-policy-adherent-tool-calling-agents/2026-06-19weekly0.7 https://korshunov.ai/en/article/1997-stylisticbias-visual-cues-drive-most-social-biases-in-mllms/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1997-stylisticbias-vizualnye-podskazki-opredeliaiut-bolshinstvo-sotsialnykh-v-mllm/2026-06-19weekly0.7 https://korshunov.ai/en/article/1998-h-replan-hierarchical-recovery-for-cross-device-agent-systems/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1998-h-replan-ierarkhicheskoe-vosstanovlenie-dlia-sistem-agentov-na-raznykh/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1996-ledgeragent-strukturirovannyi-status-dlia-agentov-vyzova-instrumentov-politiku/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1994-issledovanie-vozmozhnosti-ispolzovaniia-tesla-p40-s-uluchshennym-sistemoi/2026-06-19weekly0.7 https://korshunov.ai/en/article/1993-llama-cpp-release-b9711-new-binaries-and-updates/2026-06-19weekly0.7 https://korshunov.ai/en/article/1994-tesla-p40-feasibility-experiment-with-improved-cooling-design/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1993-llama-cpp-release-b9711-new-binaries-and-updates/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1992-llama-cpp-release-b9712-ispravliaet-sborku-interfeisa-pri-nalichii-failov-tolko/2026-06-19weekly0.7 https://korshunov.ai/en/article/1992-llama-cpp-release-b9712-fixes-ui-build-with-read-only-source/2026-06-19weekly0.7 https://korshunov.ai/en/article/1988-seti-home-as-distributed-llm-inference-engine/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1987-ai-ekonomist-agent-kadrovaia-analiticheskaia-platforma-v-oblasti-ekonomiki/2026-06-19weekly0.7 https://korshunov.ai/en/article/1987-ai-economist-agent-model-grounded-economic-analysis-framework/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1988-seti-home-kak-raspredelennaia-sistema-inferensa-llm/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1985-upravlenie-zadachami-snizhaet-zaderzhku-ocheredi-na-14-75-na-masshtabe/2026-06-19weekly0.7 https://korshunov.ai/en/article/1985-task-manager-reduces-queue-latency-by-14-75-at-enterprise-scale/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1986-see-and-reach-vision-language-navigation-for-uavs-in-field-of-view/2026-06-19weekly0.7 https://korshunov.ai/en/article/1986-see-and-reach-vision-language-navigation-for-uavs-in-field-of-view/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1984-legkii-kak-protsess-verifitsirovannyi-orakul-voznagrazhdeniia-v-rl-dlia-teorem/2026-06-19weekly0.7 https://korshunov.ai/en/article/1984-lean-as-process-verified-reward-oracle-in-rl-for-theorem-proving/2026-06-19weekly0.7 https://korshunov.ai/en/article/1983-eeg-foundation-models-for-burst-suppression-detection-in-icu/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1982-izuchaemoe-globalnoe-sliianie-dlia-peremennoi-dliny-tokenizatsii-v-diffuzii/2026-06-19weekly0.7 https://korshunov.ai/en/article/1982-learnable-global-merging-for-variable-length-tokenization-in-diffusion/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1981-skrytaia-evoliutsiia-skrytogo-vizualnogo-konteksta-v-modeliakh-s-vizualnym/2026-06-19weekly0.7 https://korshunov.ai/en/article/1981-hidden-evolution-of-disguised-visual-context-in-vlms/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1983-eeg-foundation-models-dlia-obnaruzheniia-vspleskov-podavleniia-v-otdelenii/2026-06-19weekly0.7 https://korshunov.ai/en/article/1980-residual-space-evolutionary-optimization-via-flow-based-generative-models/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1978-ihubert-monolingvalnyi-persiiskii-predobuchennyi-model-s-semanticheskoi/2026-06-19weekly0.7 https://korshunov.ai/en/article/1978-ihubert-persian-pretrained-model-with-semantic-deduplication/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1980-optimizatsiia-na-osnove-evoliutsionnykh-metodov-v-prostranstve-ostatkov-cherez/2026-06-19weekly0.7 https://korshunov.ai/en/article/1979-attention-based-sac-for-porosity-prediction-in-additive-manufacturing/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1979-vnimanie-orientirovannyi-sac-dlia-prognozirovaniia-poristosti-v-additivnom/2026-06-19weekly0.7 https://korshunov.ai/en/article/1976-hybrid-diffusion-transformer-for-instruction-guided-audio-editing/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1977-makeupmirror-uluchshaet-sokhranenie-litsevykh-kharakteristik-v-modeliakh/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1975-sensorno-motornye-modeli-mira-dlia-vospriiatiia-sinkhronizirovannogo-s/2026-06-19weekly0.7 https://korshunov.ai/en/article/1975-sensorimotor-world-models-for-action-aligned-perception/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1976-gibridnyi-diffuzionnyi-transformer-dlia-redaktirovaniia-audio-s-ispolzovaniem/2026-06-19weekly0.7 https://korshunov.ai/en/article/1977-makeupmirror-improves-facial-attribute-preservation-in-diffusion-models/2026-06-19weekly0.7 https://korshunov.ai/en/article/1973-scaffoldagent-utility-guided-dynamic-outline-optimization/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1973-scaffoldagent-utility-orientirovannyi-dinamicheskii-optimizatsiei-struktury/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1972-chastotno-osoznannoe-sovpadenie-potokov-dlia-generatsii-deistvii-robota/2026-06-19weekly0.7 https://korshunov.ai/en/article/1972-frequency-aware-flow-matching-for-robotic-action-generation/2026-06-19weekly0.7 https://korshunov.ai/en/article/1974-dual-agent-framework-for-cross-model-verified-translation/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1974-dvukhagentnaia-arkhitektura-dlia-proverennoi-perevoda-mezhdu-modeliami/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1969-bim-edit-otsenka-llm-dlia-redaktirovaniia-bim-na-osnove-ifc/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1971-adaptivnoe-obuchenie-s-pomoshchiu-llm-povyshaet-vovlechennost-i-effektivnost/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1970-racl-sloi-upravleniia-agentom-dlia-meta-optimizatsii-obucheniia/2026-06-19weekly0.7 https://korshunov.ai/en/article/1969-bim-edit-benchmarking-llms-for-ifc-based-bim-editing/2026-06-19weekly0.7 https://korshunov.ai/en/article/1970-racl-reasoning-agent-control-layer-for-metaheuristic-learning/2026-06-19weekly0.7 https://korshunov.ai/en/article/1971-adaptive-llm-tutoring-improves-engagement-and-efficiency/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1968-gibridnaia-sistema-ann-snn-s-lokalnoi-plastichnostiu/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1966-modulnaia-sparsnost-sinkhronizatsii-dlia-obucheniia-pinn/2026-06-19weekly0.7 https://korshunov.ai/en/article/1968-hybrid-ann-snn-pipeline-with-local-plasticity/2026-06-19weekly0.7 https://korshunov.ai/en/article/1966-modular-sparsity-synchronization-for-pinn-training/2026-06-19weekly0.7 https://korshunov.ai/en/article/1967-essay-quality-representations-in-llms-found-to-be-linearly-accessible/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1967-kachestvo-esse-v-bolshikh-iazykovykh-modeliakh-obnaruzhivaetsia-v-lineino/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1965-kadrovaia-sistema-semanticheskogo-rassuzhdeniia-na-osnove-gipergrafov/2026-06-19weekly0.7 https://korshunov.ai/en/article/1963-rs-neg-benchmark-and-nefo-method-for-negation-understanding-in-remote-sensing/2026-06-19weekly0.7 https://korshunov.ai/en/article/1964-medrlm-recursive-multimodal-health-intelligence-framework/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1964-medrlm-rekursivnaia-multimodalnaia-intellektualnaia-sistema-zdravookhraneniia/2026-06-19weekly0.7 https://korshunov.ai/en/article/1965-hypergraph-based-semantic-reasoning-framework/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1963-benchmark-rs-neg-i-metod-nefo-dlia-ponimaniia-otritsaniia-v-mlm-modeliakh/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1962-hilda-ierarkhicheskoe-distillirovanie-s-diffuziei-dlia-samoobucheniia-lidar/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1960-vvedenie-metriki-narusheniia-pravil-dlia-logicheskoi-soglasovannosti/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1961-psikhologicheskie-profili-llm-iavliaiutsia-artefaktami-izmereniia/2026-06-19weekly0.7 https://korshunov.ai/en/article/1961-llm-psychological-profiles-are-measurement-artifacts/2026-06-19weekly0.7 https://korshunov.ai/en/article/1960-introducing-rule-violation-score-for-logical-compliance/2026-06-19weekly0.7 https://korshunov.ai/en/article/1962-hilda-hierarchical-distillation-with-diffusion-for-self-supervised-lidar/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1957-obnaruzhenie-smeshcheniia-kontseptov-na-osnove-obuchaemykh-modelei-analiz-i/2026-06-19weekly0.7 https://korshunov.ai/en/article/1957-learner-based-concept-drift-detection-analysis-and-evaluation/2026-06-19weekly0.7 https://korshunov.ai/en/article/1958-deep-reinforcement-learning-for-game-ai-enhancement/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1958-glubokoe-obuchenie-s-podkrepleniem-dlia-uluchsheniia-ii-v-igrakh/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1959-flowmaps-modeli-dlitelnykh-mnogomodalnykh-dinamik-obektov/2026-06-19weekly0.7 https://korshunov.ai/en/article/1959-flowmaps-models-long-term-multimodal-object-dynamics/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1955-termodinamicheskaia-mera-intellekta/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1956-qmfol-otsenka-sposobnosti-bolshikh-modelei-k-logicheskomu-myshleniiu-s/2026-06-19weekly0.7 https://korshunov.ai/en/article/1956-qmfol-benchmarking-llm-reasoning-with-controllable-logical-complexity/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1954-scholarquest-baza-dlia-agentnogo-poiska-akademicheskikh-rabot-s-ispolzovaniem/2026-06-19weekly0.7 https://korshunov.ai/en/article/1954-scholarquest-taxonomy-guided-benchmark-for-agentic-academic-search/2026-06-19weekly0.7 https://korshunov.ai/en/article/1955-thermodynamic-measure-of-intelligence/2026-06-19weekly0.7 https://korshunov.ai/en/article/1951-macr-explicit-conflict-resolution-for-llm-inference/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1953-mamo-mnogoagentnaia-sistema-dlia-mnogokriterialnoi-optimizatsii-s/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1951-macr-iavnoe-razreshenie-konfliktov-v-inferense-llm/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1952-spot-e-metod-formirovaniia-entropii-na-etape-testirovaniia-s-vizualnymi-fokusami/2026-06-19weekly0.7 https://korshunov.ai/en/article/1952-spot-e-test-time-entropy-shaping-with-visual-spotlights-for-frozen-vlms/2026-06-19weekly0.7 https://korshunov.ai/en/article/1953-mamo-multi-agent-system-for-multi-objective-constrained-optimization/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1948-sootvetstvie-redaktsionnym-standartam-v-rasprostranenii-znanii-s-pomoshchiu/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1950-nastraivanie-modelei-vla-trebuet-menshe-sloev-chem-dumalos/2026-06-19weekly0.7 https://korshunov.ai/en/article/1948-editorial-alignment-in-llm-mediated-knowledge-dissemination/2026-06-19weekly0.7 https://korshunov.ai/en/article/1950-finetuning-vla-models-requires-fewer-layers-than-thought/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1949-kadrovaia-intellektualnaia-ramka-dlia-publichnoi-diskussii-v-nigerii/2026-06-19weekly0.7 https://korshunov.ai/en/article/1949-meaning-intelligence-framework-for-nigerian-public-discourse/2026-06-19weekly0.7 https://korshunov.ai/en/article/1947-confidence-aware-automated-assessment-of-student-drawn-scientific-models/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1946-lagrange-otkrytyi-slovarnyi-razrezhennyi-freimvork-dlia-tselostnogo-upravleniia/2026-06-19weekly0.7 https://korshunov.ai/en/article/1946-lagrange-open-vocabulary-sparse-framework-for-end-to-end-driving/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1947-doverie-orientirovannyi-avtomatizirovannyi-analiz-studentskikh-nauchnykh-modelei/2026-06-19weekly0.7 https://korshunov.ai/en/article/1944-boundary-embedding-shaping-for-graph-structural-disentanglement/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1943-novyi-podkhod-dtl-dlia-diagnostiki-neispravnostei-pri-ogranichennom-obeme/2026-06-19weekly0.7 https://korshunov.ai/en/article/1943-novel-dtl-approach-for-data-scarce-fault-diagnosis/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1945-elva-freimvork-na-osnove-ranzhirovaniia-dlia-mnogomodalnogo-poiska/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1944-granichnoe-vstraivanie-dlia-rasshchepleniia-struktury-grafa/2026-06-19weekly0.7 https://korshunov.ai/en/article/1945-elva-a-ranking-driven-framework-for-multimodal-retrieval/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1942-softskill-szhatie-povedencheskikh-navykov-dlia-adaptatsii-v-kontekste/2026-06-19weekly0.7 https://korshunov.ai/en/article/1940-trajectory-mining-reveals-skill-structure-but-fails-to-improve-policies/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1940-analiz-traektorii-raskryvaet-strukturu-navykov-no-ne-uluchshaet-politiki/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1941-ustoichivoe-obuchenie-q-dlia-upravleniia-polem-sredy-pri-neopredelennosti/2026-06-19weekly0.7 https://korshunov.ai/en/article/1941-robust-q-learning-for-mean-field-control-under-wasserstein-uncertainty/2026-06-19weekly0.7 https://korshunov.ai/en/article/1942-softskill-behavioral-compression-for-contextual-adaptation/2026-06-19weekly0.7 https://korshunov.ai/en/article/1938-crax-fast-safe-reinforcement-learning-benchmarking/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1939-autopass-agenty-na-osnove-dokazatelstv-dlia-nastroiki-proizvoditelnosti/2026-06-19weekly0.7 https://korshunov.ai/en/article/1939-autopass-evidence-guided-llm-agents-for-compiler-performance-tuning/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1938-crax-bystryi-bezopasnyi-benchmark-dlia-obucheniia-s-usileniem/2026-06-19weekly0.7 https://korshunov.ai/en/article/1934-streamkl-fast-and-memory-efficient-kl-divergence-for-attention-distillation/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1934-streamkl-bystryi-i-pamiat-effektivnyi-kl-razbros-dlia-distilliatsii-vnimaniia/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1936-tri-info-obshchee-predskazanie-sboev-dlia-modelei-vla/2026-06-19weekly0.7 https://korshunov.ai/en/article/1936-tri-info-generalizable-failure-prediction-for-vla-models/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1935-obuchenie-bolshikh-iazykovykh-modelei-dlia-agentov-s-dlitelnym-zhiznennym-cherez/2026-06-19weekly0.7 https://korshunov.ai/en/article/1935-training-llms-for-long-lifecycle-agents-via-cross-domain-generalization/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1932-samoadaptivnaia-obrabotka-masshtabov-dlia-prognozirovaniia-vremennykh-riadov/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1933-vimpo-kriterii-besplatnaia-optimizatsiia-politiki-dlia-bolshikh-iazykovykh/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1931-kontrol-na-osnove-llm-v-mnogokontroliruemykh-igrakh/2026-06-19weekly0.7 https://korshunov.ai/en/article/1933-vimpo-critic-free-policy-optimization-for-llms/2026-06-19weekly0.7 https://korshunov.ai/en/article/1931-llm-based-hierarchical-control-in-multi-agent-games/2026-06-19weekly0.7 https://korshunov.ai/en/article/1932-self-adaptive-scale-handling-for-time-series-forecasting/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1928-tessera-i-alphaearth-vstraivaniia-pozvoliaiut-provodit-melkoe-kartirovanie-lcz-v/2026-06-19weekly0.7 https://korshunov.ai/en/article/1928-tessera-and-alphaearth-embeddings-enable-fine-scale-lcz-mapping-in-swiss-cities/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1930-ad-deeponet-dlia-bystrogo-prognozirovaniia-otveta-mosta/2026-06-19weekly0.7 https://korshunov.ai/en/article/1930-ad-deeponet-for-fast-bridge-response-prediction/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1929-sme-ofu-podkhod-na-osnove-mnozhestva-chlenov-dlia-stokhasticheskikh-lineinykh/2026-06-19weekly0.7 https://korshunov.ai/en/article/1929-sme-ofu-set-membership-approach-for-stochastic-linear-contextual-bandits/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1926-diagnostika-altsgeimera-s-ispolzovaniem-sliianiia-multimodalnykh-3d-mrt-i-pet/2026-06-19weekly0.7 https://korshunov.ai/en/article/1925-ai-economist-agent-model-grounded-economic-analysis-framework/2026-06-19weekly0.7 https://korshunov.ai/en/article/1926-alzheimer-s-diagnosis-via-multimodal-3d-mri-and-pet-fusion/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1927-pu-unet-stabilnye-multiplikativnye-vzaimodeistviia-dlia-segmentatsii/2026-06-19weekly0.7 https://korshunov.ai/en/article/1927-pu-unet-stable-multiplicative-interactions-for-medical-image-segmentation/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1925-ai-ekonomist-agent-freimvork-analiza-na-osnove-modelei/2026-06-19weekly0.7 https://korshunov.ai/en/article/1923-paano-lightweight-time-series-anomaly-detection-with-multiscale-and-cross/2026-06-19weekly0.7 https://korshunov.ai/en/article/1922-optimal-coarse-correlated-equilibria-in-mean-field-games/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1923-paano-legkaia-model-obnaruzheniia-anomalii-v-vremennykh-riadakh-s-i-vnimaniem/2026-06-19weekly0.7 https://korshunov.ai/en/article/1924-comparative-study-of-neural-surrogates-for-battery-state-prediction/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1924-sravnitelnoe-issledovanie-neironnykh-zamenitelei-dlia-prognoza-sostoianiia/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1922-optimalnye-koardinatsionnye-korrelirovannye-ravnovesiia-v-igrakh-s-polem-sredy/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1919-razdelenie-zatrat-v-metode-dzhonsa-ellipsoida/2026-06-19weekly0.7 https://korshunov.ai/en/article/1920-information-theoretic-analysis-of-effective-supervision-in-latent-chain-of/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1920-informatsionnyi-analiz-effektivnogo-nadzora-v-skrytoi-tsepochke-myslei/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1921-eeg-foundation-models-dlia-obnaruzheniia-vspleskov-podavleniia-v-icu/2026-06-19weekly0.7 https://korshunov.ai/en/article/1921-eeg-foundation-models-for-burst-suppression-detection-in-icu/2026-06-19weekly0.7 https://korshunov.ai/en/article/1919-decoupling-costs-in-john-ellipsoid-approximation/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1918-makeupmirror-uluchshaet-sokhranenie-atributov-litsa-v-modeliakh-diffuzii/2026-06-19weekly0.7 https://korshunov.ai/en/article/1916-quantile-of-means-ensemble-method-for-minimax-optimal-rl/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1916-kvantil-srednikh-metod-ansamblia-dlia-minimaks-optimalnogo-obucheniia-s/2026-06-19weekly0.7 https://korshunov.ai/en/article/1917-sensorimotor-world-models-for-action-aligned-perception/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1917-sensorno-motornye-modeli-mira-dlia-vospriiatiia-sinkhronizirovannogo-s/2026-06-19weekly0.7 https://korshunov.ai/en/article/1918-makeupmirror-improves-facial-attribute-preservation-in-diffusion-models/2026-06-19weekly0.7 https://korshunov.ai/en/article/1913-pose6daug-physically-plausible-multi-view-object-swapping/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1913-pose6daug-fizicheski-obosnovannyi-obmen-obektami-v-mnogokamernom-vide/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1915-efiqa-otsenka-kachestva-fundus-izobrazhenii-bez-metok-s-vozmozhnostiu/2026-06-19weekly0.7 https://korshunov.ai/en/article/1914-federated-conformal-risk-control-via-risk-curve-shrinkage/2026-06-19weekly0.7 https://korshunov.ai/en/article/1915-efiqa-label-free-fundus-image-quality-assessment-with-explainability/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1914-federativnoe-upravlenie-riskom-cherez-szhatie-krivoi-riska/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1912-generiruemye-llm-iadra-gpu-podvergaiutsia-illiuzii-korrektnosti/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1910-adaptivnoe-obuchenie-s-ispolzovaniem-llm-povyshaet-vovlechennost-i-effektivnost/2026-06-19weekly0.7 https://korshunov.ai/en/article/1910-adaptive-llm-tutoring-improves-engagement-and-efficiency/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1911-pasqa-model-otsenki-kachestva-rechi-s-fokusom-na-intonatsionnye-aktsenty/2026-06-19weekly0.7 https://korshunov.ai/en/article/1911-pasqa-pitch-accent-focused-speech-quality-model/2026-06-19weekly0.7 https://korshunov.ai/en/article/1912-llm-generated-gpu-kernels-face-correctness-illusion/2026-06-19weekly0.7 https://korshunov.ai/en/article/1909-medrlm-recursive-multimodal-health-intelligence-framework/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1907-mashinnoe-obuchenie-predskazyvaet-vozrast-beremennosti-po-magnitno-rezonansnoi/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1908-melt-i-salt-multimodalnye-modeli-kontrastnogo-obucheniia-dlia-zemnykh-vlozhenii/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1909-medrlm-rekursivnaia-multimodalnaia-intellektualnaia-sistema-v-oblasti/2026-06-19weekly0.7 https://korshunov.ai/en/article/1907-machine-learning-predicts-gestational-age-from-fetal-mri/2026-06-19weekly0.7 https://korshunov.ai/en/article/1908-melt-and-salt-multimodal-contrastive-learning-for-earth-embeddings/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1906-vychislitelnye-metody-dlia-dnk-vne-kletok-v-rannem-vyiavlenii-mnozhestvennykh/2026-06-19weekly0.7 https://korshunov.ai/en/article/1905-effective-dimension-governs-generalization-in-quantum-vision-models/2026-06-19weekly0.7 https://korshunov.ai/en/article/1906-computational-methods-for-cell-free-dna-in-multi-cancer-early-detection/2026-06-19weekly0.7 https://korshunov.ai/en/article/1904-off-policy-evaluation-for-mnar-rewards-in-mdps/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1904-off-politicheskii-analiz-dlia-voznagrazhdenii-mnar-v-mdp/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1905-effektivnaia-razmernost-opredeliaet-obobshchenie-v-kvantovykh-modeliakh-zreniia/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1901-granichnoe-vstraivanie-i-formirovanie-dlia-rasshchepleniia-struktur-grafa/2026-06-19weekly0.7 https://korshunov.ai/en/article/1903-learner-based-concept-drift-detection-analysis-and-evaluation/2026-06-19weekly0.7 https://korshunov.ai/en/article/1901-boundary-embedding-shaping-for-graph-structural-disentanglement/2026-06-19weekly0.7 https://korshunov.ai/en/article/1902-mamo-multi-agent-system-for-multi-objective-constrained-optimization/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1903-obnaruzhenie-smeshcheniia-kontseptsii-na-osnove-obuchaemykh-modelei-analiz-i/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1902-mamo-mnogoagentnaia-sistema-dlia-mnogokriterialnoi-optimizatsii-s/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1900-kadr-vibrantforests-otobrazhaet-strukturu-lesov-s-razresheniem-10-metrov/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1898-statisticheskie-svoistva-obucheniia-i-obobshcheniia/2026-06-19weekly0.7 https://korshunov.ai/en/article/1899-slir-shifting-based-linear-relaxations-for-activation-functions/2026-06-19weekly0.7 https://korshunov.ai/en/article/1898-statistical-properties-of-training-and-generalization/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1899-slir-lineinye-relaksatsii-na-osnove-sdviga-dlia-funktsii-aktivatsii/2026-06-19weekly0.7 https://korshunov.ai/en/article/1900-vibrantforests-framework-maps-forest-structure-at-10-meter-resolution/2026-06-19weekly0.7 https://korshunov.ai/en/article/1895-qcpikan-quantum-classical-physics-informed-kan-for-pdes/2026-06-19weekly0.7 https://korshunov.ai/en/article/1897-model-driven-approach-for-rl-environment-families/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1896-rekurrentnye-neironnye-seti-approksimiruiut-nepreryvnye-funktsii/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1895-qcpikan-kvantovo-klassicheskaia-fizicheski-obuslovlennaia-set-kolmogorova-dlia-v/2026-06-19weekly0.7 https://korshunov.ai/en/article/1896-recurrent-neural-networks-approximate-continuous-functions/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1897-modelnyi-podkhod-k-semeistvam-sred-obucheniia-s-usileniem/2026-06-19weekly0.7 https://korshunov.ai/en/article/1894-hybrid-modeling-predicts-microbial-dynamics-in-soil-systems/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1892-kriticheskaia-perkoliatsiia-kak-model-sinteticheskikh-dannykh-dlia/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1894-gibridnaia-model-predskazyvaet-dinamiku-mikroorganizmov-v-pochvennykh-sistemakh/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1893-kvantovyi-koltsevoi-agregat-preimushchestva-v-kommunikatsii-i-privatnosti-dlia/2026-06-19weekly0.7 https://korshunov.ai/en/article/1893-quantum-ring-all-reduce-communication-and-privacy-advantages-for-distributed/2026-06-19weekly0.7 https://korshunov.ai/en/article/1892-critical-percolation-as-a-synthetic-data-model-for-interpretability/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1891-ustoichivoe-obuchenie-q-dlia-upravleniia-polem-srednego-pri-vodianom/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1889-obuchenie-izvlechenie-ili-oba-varianta-priamoe-sravnenie-po-statutarnoi-tsitate/2026-06-19weekly0.7 https://korshunov.ai/en/article/1889-train-retrieve-or-both-head-to-head-on-statutory-citation-for-ontario-rta/2026-06-19weekly0.7 https://korshunov.ai/en/article/1891-robust-q-learning-for-mean-field-control-under-wasserstein-uncertainty/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1890-snizhenie-dispersii-v-obuchenii-na-osnove-vremennykh-raznostei/2026-06-19weekly0.7 https://korshunov.ai/en/article/1890-variance-reduction-in-temporal-difference-learning/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1886-primechaniia-k-vypusku-v2-1-183/2026-06-19weekly0.7 https://korshunov.ai/en/article/1888-de-biased-vlm-as-3d-judge-protocol-for-furniture-generation/2026-06-19weekly0.7 https://korshunov.ai/en/article/1886-v2-1-183-release-notes/2026-06-19weekly0.7 https://korshunov.ai/en/article/1887-crax-fast-safe-reinforcement-learning-benchmarking/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1888-protokol-de-biasirovannogo-vlm-kak-3d-sudi-dlia-generatsii-mebeli/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1887-crax-bystraia-i-bezopasnaia-otsenka-v-oblasti-obucheniia-s-usileniem/2026-06-19weekly0.7 https://korshunov.ai/en/article/1885-cot-transformers-can-efficiently-simulate-word-ram-algorithms/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1884-analiz-nastroeniia-propuskaet-kliuchevye-rezultaty-klientov/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1883-terramars-paiplain-malogo-iazykovogo-modelirovaniia-dlia-literatury-po-marsa/2026-06-19weekly0.7 https://korshunov.ai/en/article/1884-sentiment-analysis-misses-key-customer-outcomes/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1885-transformatory-s-tsepochkoi-myslei-mogut-effektivno-imitirovat-algoritmy-word/2026-06-19weekly0.7 https://korshunov.ai/en/article/1883-terramars-small-language-model-pipeline-for-mars-terraforming-literature/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1882-nest-dataset-dlia-narrativnykh-struktur-sobytii-v-dlinnykh-video/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1880-vvedenie-p-chr-auc-i-crr-dlia-semanticheskoi-keshirovaniia/2026-06-19weekly0.7 https://korshunov.ai/en/article/1881-finerex-fine-tuned-ner-re-for-human-smuggling-knowledge-graphs/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1881-finerex-nastroennaia-sistema-izvlecheniia-nazvanii-i-sviazei-dlia-znanii-o/2026-06-19weekly0.7 https://korshunov.ai/en/article/1880-introducing-p-chr-auc-and-crr-for-semantic-caching/2026-06-19weekly0.7 https://korshunov.ai/en/article/1882-nest-dataset-for-narrative-event-structures-in-long-videos/2026-06-19weekly0.7 https://korshunov.ai/en/article/1877-benchmarking-agentic-review-systems-for-ai-assisted-research/2026-06-19weekly0.7 https://korshunov.ai/en/article/1878-sequential-dpo-shows-variable-preference-impact-across-settings/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1877-otsenka-agentnykh-sistem-analiza-dlia-issledovanii-s-uchastiem-ii/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1879-nrityam-benchmark-dlia-kulturnogo-ponimaniia-v-tantse/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1878-posledovatelnyi-dpo-pokazyvaet-peremennoe-vliianie-predpochtenii-v-razlichnykh/2026-06-19weekly0.7 https://korshunov.ai/en/article/1879-nrityam-benchmark-for-cultural-comprehension-in-dance/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1874-combeval-benchmark-dlia-podscheta-kombinatornykh-zadach-v-llms/2026-06-19weekly0.7 https://korshunov.ai/en/article/1874-combeval-benchmark-for-combinatorial-counting-in-llms/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1876-baiesovskoe-kurrikuliarnoe-obuchenie-na-latentnykh-mnogoobraziiakh-llm/2026-06-19weekly0.7 https://korshunov.ai/en/article/1876-bayesian-curriculum-learning-on-llm-latent-manifolds/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1875-agentfinvqa-auditiruemyi-lokalnyi-vopros-otvet-po-finansovym-grafikam/2026-06-19weekly0.7 https://korshunov.ai/en/article/1875-agentfinvqa-auditable-on-premise-financial-chart-qa/2026-06-19weekly0.7 https://korshunov.ai/en/article/1872-semantic-clusters-pre-train-tsetlin-machine-for-interpretability/2026-06-19weekly0.7 https://korshunov.ai/en/article/1871-credence-semantic-metrics-and-convergence-analysis-for-claim-decomposition/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1871-credence-semanticheskie-metriki-i-analiz-skhodimosti-dlia-razbieniia/2026-06-19weekly0.7 https://korshunov.ai/en/article/1873-selective-verification-for-budget-aware-reasoning/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1873-selektivnaia-proverka-dlia-rassuzhdenii-s-uchetom-biudzheta/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1872-semanticheskie-klastery-predobuchenie-tetslin-mashiny-dlia-interpretiruemosti/2026-06-19weekly0.7 https://korshunov.ai/en/article/1870-jamer-project-level-code-framework-dataset-and-benchmark/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1868-atommem-prostaia-i-effektivnaia-sistema-pamiati-dlia-agentov-llm/2026-06-19weekly0.7 https://korshunov.ai/en/article/1868-atommem-simple-and-effective-memory-system-for-llm-agents/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1870-jamer-dataset-i-benchmark-proektnogo-urovnia-koda/2026-06-19weekly0.7 https://korshunov.ai/en/article/1869-control-window-law-for-single-neuron-steering-in-language-models/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1869-zakon-kontrolia-okna-dlia-upravleniia-odnim-neironom-v-iazykovykh-modeliakh/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1867-nulevaia-zadacha-agentnykh-llm-izvlekaet-patologiiu-legkikh-iz-povestvovanii/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1865-ai-driven-deliberation-scaling-inclusivity-and-empowering-marginalised-groups/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1866-llms-mogut-obrabatyvat-neprochityvaemyi-tekst-s-vysokoi-semanticheskoi/2026-06-19weekly0.7 https://korshunov.ai/en/article/1865-ai-driven-deliberation-scaling-inclusivity-and-empowering-marginalised-groups/2026-06-19weekly0.7 https://korshunov.ai/en/article/1867-zero-shot-agentic-llms-extract-lung-pathology-from-narratives/2026-06-19weekly0.7 https://korshunov.ai/en/article/1866-llms-can-process-non-readable-text-with-high-semantic-fidelity/2026-06-19weekly0.7 https://korshunov.ai/en/article/1863-lightweight-pronunciation-assessment-via-discrete-speech-token-surprisal/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1862-freimvork-multiagentnoi-tranzaktsionnoi-pamiati/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1864-redact-multimaliarnyi-benchmark-po-obnaruzheniiu-personalnykh-dannykh-s/2026-06-19weekly0.7 https://korshunov.ai/en/article/1864-redact-multilingual-pii-benchmark-with-systematic-control/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1863-legkaia-otsenka-proiznosheniia-s-pomoshchiu-nesviazannogo-spektralnogo-siurpriza/2026-06-19weekly0.7 https://korshunov.ai/en/article/1862-multi-agent-transactive-memory-framework/2026-06-19weekly0.7 https://korshunov.ai/en/article/1858-training-llms-for-long-lifecycle-agents-via-cross-domain-generalization/2026-06-19weekly0.7 https://korshunov.ai/en/article/1860-speech-quality-models-fail-to-capture-prosodic-and-f0-variability/2026-06-19weekly0.7 https://korshunov.ai/en/article/1861-gems-geometric-constraints-enable-multi-semantic-superposition-in-llms/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1861-gems-geometricheskie-ogranicheniia-obespechivaiut-multi-semanticheskuiu-v-llms/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1858-obuchenie-bolshikh-iazykovykh-modelei-dlia-agentov-s-dlitelnym-zhiznennym-cherez/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1860-modeli-kachestva-rechi-ne-ulavlivaiut-variativnost-pronutsiatsii-i-chastoty/2026-06-19weekly0.7 https://korshunov.ai/en/article/1859-segment-level-mandarin-speech-detection-for-cognitive-impairment/2026-06-19weekly0.7 https://korshunov.ai/en/article/1857-over-privileged-tool-selection-in-llm-agents/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1859-obnaruzhenie-rechi-na-urovne-segmentov-dlia-vyiavleniia-kognitivnykh-narushenii/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1857-pereotsenivanie-privilegii-pri-vybore-instrumentov-v-agentakh-na-osnove-modelei/2026-06-19weekly0.7 https://korshunov.ai/en/article/1855-stage-source-grounded-data-generation-for-text-to-json/2026-06-19weekly0.7 https://korshunov.ai/en/article/1854-information-theoretic-analysis-of-effective-supervision-in-latent-chain-of/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1854-informatsionnyi-analiz-effektivnogo-nadzora-v-skrytoi-tsepochke-myslei/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1855-stage-generatsiia-dannykh-s-oporoi-na-istochnik-dlia-preobrazovaniia-teksta-v/2026-06-19weekly0.7 https://korshunov.ai/en/article/1856-generative-engine-optimization-measuring-ai-search-visibility/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1856-optimizatsiia-generativnoi-sistemy-izmerenie-vidimosti-ai-v-poiske/2026-06-19weekly0.7 https://korshunov.ai/en/article/1851-hydrahead-head-level-hybrid-attention-for-long-context-performance/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1851-hydrahead-gibridnoe-vnimanie-na-urovne-golovok-dlia-proizvoditelnosti-pri/2026-06-19weekly0.7 https://korshunov.ai/en/article/1852-no-self-preference-in-model-revision-under-genuine-authorship/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1853-ihubert-monolingvalnyi-predobuchennyi-model-na-persidskom-iazyke-s/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1852-otsutstvie-samopredpochteniia-pri-redaktirovanii-modelei-pri-nastoiashchem/2026-06-19weekly0.7 https://korshunov.ai/en/article/1853-ihubert-persian-pretrained-model-with-semantic-deduplication/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1849-pasqa-model-otsenki-kachestva-rechi-s-fokusom-na-intonatsionnykh-aktsentakh/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1850-stabilizatsiia-namereniia-instrumenta-v-potokovom-rag/2026-06-19weekly0.7 https://korshunov.ai/en/article/1850-tool-intent-stabilization-in-streaming-rag/2026-06-19weekly0.7 https://korshunov.ai/en/article/1849-pasqa-pitch-accent-focused-speech-quality-model/2026-06-19weekly0.7 https://korshunov.ai/en/article/1848-adaptive-llm-tutoring-improves-engagement-and-efficiency/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1848-adaptivnoe-obuchenie-s-pomoshchiu-llm-povyshaet-vovlechennost-i-effektivnost/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1846-chernyi-iashchik-obnaruzhivaet-zapominanie-identichnostei-v-modeliakh-tekst-v/2026-06-19weekly0.7 https://korshunov.ai/en/article/1845-medrlm-recursive-multimodal-health-intelligence-framework/2026-06-19weekly0.7 https://korshunov.ai/en/article/1847-essay-quality-representations-in-llms-found-to-be-linearly-accessible/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1847-kachestvo-esse-v-bolshikh-iazykovykh-modeliakh-obnaruzhivaetsia-v-lineino/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1845-medrlm-rekursivnaia-multimodalnaia-intellektualnaia-sistema-zdravookhraneniia/2026-06-19weekly0.7 https://korshunov.ai/en/article/1846-black-box-probe-detects-identity-memorization-in-text-to-image-models/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1843-algoritm-dlia-opredeleniia-not-i-kliuchei-v-muzykalnoi-transkriptsii/2026-06-19weekly0.7 https://korshunov.ai/en/article/1843-algorithm-for-pitch-spelling-and-key-estimation-in-music-transcription/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1844-renikud-audio-obuchaemaia-konvertatsiia-grafemy-v-zvuk-v-ivrite/2026-06-19weekly0.7 https://korshunov.ai/en/article/1844-renikud-audio-supervised-hebrew-grapheme-to-phoneme-conversion/2026-06-19weekly0.7 https://korshunov.ai/en/article/1842-llm-psychological-profiles-are-measurement-artifacts/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1842-psikhologicheskie-profili-bolshikh-iazykovykh-modelei-eto-artefakty-izmereniia/2026-06-19weekly0.7 https://korshunov.ai/en/article/1839-meaning-intelligence-framework-for-nigerian-public-discourse/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1841-czechdocs-parallelnyi-nabor-dannykh-dlia-perevoda-dokumentov-na-malye-iazyki/2026-06-19weekly0.7 https://korshunov.ai/en/article/1840-causal-activation-directions-for-mitigating-emergent-misalignment-in-language/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1840-kauzalnye-napravleniia-aktivatsii-dlia-smiagcheniia-emerdzhentnoi-v-iazykovykh/2026-06-19weekly0.7 https://korshunov.ai/en/article/1841-czechdocs-parallel-dataset-for-minority-language-document-translation/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1839-kadrovaia-intellektualnaia-ramka-dlia-publichnogo-diskursa-v-nigerii/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1837-tekhniki-optimizatsii-inferensa-orientirovannye-na-tokeny/2026-06-19weekly0.7 https://korshunov.ai/en/article/1838-psyscore-a-psychometrically-aware-framework-for-trait-adaptive-essay-scoring-and/2026-06-19weekly0.7 https://korshunov.ai/en/article/1837-token-operations-oriented-inference-optimization-techniques/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1838-psyscore-freimvork-orientirovannyi-na-psikhometricheskie-aspekty-dlia-otsenki-s/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1836-catch-me-esli-vy-rag-multimediinyi-dataset-dlia-protivodeistviia-nenavisti-i/2026-06-19weekly0.7 https://korshunov.ai/en/article/1836-catch-me-if-you-rag-multilingual-counterspeech-dataset-for-hate-and/2026-06-19weekly0.7 https://korshunov.ai/en/article/1832-llama-cpp-release-b9707-adds-schema-validation-and-new-binary-builds/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1832-llama-cpp-release-b9707-dobavliaet-proverku-skhemy-i-novye-binarnye-sborki/2026-06-19weekly0.7 https://korshunov.ai/en/article/1828-datasette-apps-0-1a2-release-notes/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1831-glm-5-2-prevoskhodit-gpt-5-5-v-otsenke-aa-briefcase/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1828-datasette-apps-0-1a2-release-notes/2026-06-19weekly0.7 https://korshunov.ai/en/article/1831-glm-5-2-outperforms-gpt-5-5-in-aa-briefcase-evaluation/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1827-datasette-apps-0-1a3-release/2026-06-19weekly0.7 https://korshunov.ai/en/article/1825-datasette-launches-apps-plugin-for-custom-html-applications/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1825-datasette-predstavliaet-plagin-apps-dlia-polzovatelskikh-prilozhenii-na-html/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1826-datasette-acl-0-6a0-release/2026-06-19weekly0.7 https://korshunov.ai/en/article/1827-datasette-apps-0-1a3-release/2026-06-19weekly0.7 https://korshunov.ai/en/article/1826-datasette-acl-0-6a0-release/2026-06-19weekly0.7 https://korshunov.ai/en/article/1829-glm-5-2-744b-2-bit-achieves-7-3-tok-s-on-4x3090-with-192gb-ram/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1829-glm-5-2-744b-2-bit-dostigaet-73-tok-s-na-4x3090-s-192-gb-ozu/2026-06-19weekly0.7 https://korshunov.ai/ru/article/1823-lq50-lq50-24gb-stoit-okolo-1200/2026-06-18weekly0.7 https://korshunov.ai/en/article/1823-lq50-lq50-24gb-costs-around-1200/2026-06-18weekly0.7 https://korshunov.ai/en/article/1821-diffusiongemma-26b-on-4090-reaches-475t-s-with-limitations/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1821-diffusiongemma-26b-na-4090-dostigaet-475t-s-s-ogranicheniiami/2026-06-18weekly0.7 https://korshunov.ai/en/article/1818-what-s-the-best-open-speech-to-text-today/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1817-zapusk-glm-5-2-tolko-na-cpu-s-lokalnoi-ustanovkoi/2026-06-18weekly0.7 https://korshunov.ai/en/article/1817-running-glm-5-2-on-cpu-only-with-local-setup/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1818-kakoi-samyi-luchshii-otkrytyi-instrument-dlia-preobrazovaniia-rechi-v-tekst/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1813-langgraph-vypuskaet-versiiu-1-2-6/2026-06-18weekly0.7 https://korshunov.ai/en/article/1813-langgraph-releases-version-1-2-6/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1808-primechaniia-k-vypusku-v1-14-8a1/2026-06-18weekly0.7 https://korshunov.ai/en/article/1808-v1-14-8a1-release-notes/2026-06-18weekly0.7 https://korshunov.ai/en/article/1809-repurposing-an-old-multi-gpu-node-for-local-inference/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1809-pereispolzovanie-starogo-uzla-s-neskolkimi-gpu-dlia-lokalnoi-inferentsii/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1806-lokalnaia-qwen-ne-khuzhe-opus-eto-instrument-s-drugimi-vozmozhnostiami/2026-06-18weekly0.7 https://korshunov.ai/en/article/1806-local-qwen-isn-t-a-worse-opus-it-s-a-different-tool/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1797-kalibrovka-2-bitnykh-gguf-dlia-zadach-agentnogo-programmirovaniia/2026-06-18weekly0.7 https://korshunov.ai/en/article/1797-calibrating-2-bit-ggufs-for-agentic-coding-tasks/2026-06-18weekly0.7 https://korshunov.ai/en/article/1794-why-ai-scaling-is-a-systems-problem-not-just-a-gpu-race/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1794-pochemu-masshtabirovanie-ii-eto-problema-sistem-a-ne-prosto-sorevnovanie-po-gpu/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1793-north-mini-code-4-bitnaia-kvantovanie-podderzhka-ollama-i-openrouter/2026-06-18weekly0.7 https://korshunov.ai/en/article/1793-north-mini-code-4-bit-quant-ollama-and-openrouter-support/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1792-lfm2-5-embedding-35-0m-i-lfm2-5-colbert-350m-byli-vypushcheny/2026-06-18weekly0.7 https://korshunov.ai/en/article/1792-lfm2-5-embedding-350m-and-lfm2-5-colbert-350m-released/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1786-snizhenie-zatrat-na-tokeny-v-realnykh-usloviiakh-za-schet-headroom-rtk-i-caveman/2026-06-18weekly0.7 https://korshunov.ai/en/article/1786-real-world-token-cost-savings-from-rtk-headroom-and-caveman/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1785-laguna-m-1-225b-parametr-model-moe-dlia-agentnogo-kodirovaniia/2026-06-18weekly0.7 https://korshunov.ai/en/article/1785-laguna-m-1-225b-parameter-moe-model-for-agentic-coding/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1783-llama-cpp-release-b9703-obnovleniia-i-binarnye-zagruzki/2026-06-18weekly0.7 https://korshunov.ai/en/article/1783-llama-cpp-release-b9703-updates-and-binary-downloads/2026-06-18weekly0.7 https://korshunov.ai/en/article/1781-my-suitcase-robot-gets-high-from-real-gas-sensor/2026-06-18weekly0.7 https://korshunov.ai/en/article/1782-llama-cpp-release-b9704-fixes-invalid-grammar-handling-and-adds-new-binaries/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1781-moi-robot-sumka-poluchaet-effekt-ot-realnogo-datchika-gaza/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1782-llama-cpp-release-b9704-fixes-invalid-grammar-handling-and-adds-new-binaries/2026-06-18weekly0.7 https://korshunov.ai/en/article/1777-mistral-rs-v0-8-10-adds-v1-skills-support-for-local-models/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1777-mistral-rs-v0-8-10-dobavliaet-podderzhku-v1-skills-dlia-lokalnykh-modelei/2026-06-18weekly0.7 https://korshunov.ai/en/article/1776-glm-5-2-inference-free-on-hugging-face-for-next-6-hours/2026-06-18weekly0.7 https://korshunov.ai/en/article/1772-unsloth-glm-5-2-gguf-with-2bit-quantization-at-238gb/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1772-unsloth-glm-5-2-gguf-s-2-bitnoi-kvantovaniei-na-238-gb/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1776-besplatnyi-dostup-k-inferensu-glm-5-2-na-hugging-face-v-techenie-sleduiushchikh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1773-glm-5-2-is-the-best-open-weight-creative-writing-model/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1773-glm-5-2-luchshaia-otkrytaia-model-kreativnogo-pisma/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1768-llama-cpp-release-b9702-ispravleniia-i-novye-binarniki/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1767-luchshee-mesto-dlia-prodazhi-pochti-neispolzovannogo-rtx-pro-6000-blackwell-max/2026-06-18weekly0.7 https://korshunov.ai/en/article/1768-llama-cpp-release-b9702-fixes-and-new-binaries/2026-06-18weekly0.7 https://korshunov.ai/en/article/1767-best-place-to-sell-a-barely-used-rtx-pro-6000-blackwell-max-q/2026-06-18weekly0.7 https://korshunov.ai/en/article/1762-white-house-pauses-ai-deployment/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1760-kvantovanie-kesha-kv-nvfp4-na-sm120-sdelaet-sistemy-s-32-gb-ozu-ochen-moshchnymi/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1762-belyi-dom-ostanavlivaet-vnedrenie-ii/2026-06-18weekly0.7 https://korshunov.ai/en/article/1761-slms-and-diffusion-the-future-of-small-specialized-models/2026-06-18weekly0.7 https://korshunov.ai/en/article/1760-nvfp4-kv-cache-quantization-on-sm120-will-make-32gb-vram-systems-very-capable/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1761-slms-i-diffuzii-budushchee-malykh-spetsializirovannykh-modelei/2026-06-18weekly0.7 https://korshunov.ai/en/article/1758-glm-s-founder-says-glm-fable-before-the-end-of-the-year/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1759-moshch-intellekta-luchshe-lezhit-v-rukakh-liudei-chem-v-kabinetakh-magnatov/2026-06-18weekly0.7 https://korshunov.ai/en/article/1759-the-power-of-intelligence-is-better-in-the-hands-of-the-people-than-in-the-board/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1757-llama-cpp-release-b9701-new-preprocessor-and-cross-platform-binaries/2026-06-18weekly0.7 https://korshunov.ai/en/article/1757-llama-cpp-release-b9701-new-preprocessor-and-cross-platform-binaries/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1758-sozdatel-glm-govorit-chto-glm-fable-mozhet-byt-vypushchen-k-kontsu-goda/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1756-open-sous-modeli-decisively-overtook-proprietary-models-in-market-share/2026-06-18weekly0.7 https://korshunov.ai/en/article/1756-oss-models-decisively-overtook-proprietary-models-in-market-share/2026-06-18weekly0.7 https://korshunov.ai/en/article/1746-llama-bench-vs-real-world-performance-discrepancy/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1746-llama-bench-vs-realnoe-otklonenie-proizvoditelnosti/2026-06-18weekly0.7 https://korshunov.ai/en/article/1744-local-llm-agent-now-generates-images-and-video-offline/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1744-lokalnyi-agent-llm-teper-generiruet-izobrazheniia-i-video-v-avtonomnom-rezhime/2026-06-18weekly0.7 https://korshunov.ai/en/article/1741-llama-cpp-release-b9700-new-binaries-and-sycl-api-updates/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1741-llama-cpp-release-b9700-new-binaries-and-sycl-api-updates/2026-06-18weekly0.7 https://korshunov.ai/en/article/1738-llama-cpp-release-b9698-adds-self-update-support-and-multiple-platform-binaries/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1739-dostupnost-lokalnoi-versii-le-chaton-fat-flash/2026-06-18weekly0.7 https://korshunov.ai/en/article/1739-le-chaton-fat-flash-local-version-availability/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1738-llama-cpp-release-b9698-dobavliaet-podderzhku-samopodpiski-i-binarniki-dlia/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1737-llama-cpp-release-b9699-adds-sycl-support-and-multiple-platform-binaries/2026-06-18weekly0.7 https://korshunov.ai/en/article/1737-llama-cpp-release-b9699-adds-sycl-support-and-multiple-platform-binaries/2026-06-18weekly0.7 https://korshunov.ai/en/article/1735-keye-vl-2-0-30b-a3b-launches-with-advanced-video-understanding-and-agent/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1735-zapusk-keye-vl-2-0-30b-a3b-s-prodvinutymi-vozmozhnostiami-ponimaniia-video-i/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1736-ml-models-recommended-for-m5-max-macbook-pro-with-128gb-ram/2026-06-18weekly0.7 https://korshunov.ai/en/article/1736-ml-models-recommended-for-m5-max-macbook-pro-with-128gb-ram/2026-06-18weekly0.7 https://korshunov.ai/en/article/1729-glm-5-2-flash-release-date-joke/2026-06-18weekly0.7 https://korshunov.ai/en/article/1730-llama-cpp-release-b9697-new-binaries-and-updates/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1729-data-bystroi-versii-glm-5-2-shutka/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1730-reliz-llama-cpp-b9697-novye-binarniki-i-obnovleniia/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1726-reliz-autogpt-platform-beta-v0-6-64/2026-06-18weekly0.7 https://korshunov.ai/en/article/1726-autogpt-platform-beta-v0-6-64-released/2026-06-18weekly0.7 https://korshunov.ai/en/article/1724-does-anyone-have-enough-compute-to-make-a-distillation-dataset-from-glm5-2/2026-06-18weekly0.7 https://korshunov.ai/en/article/1723-llama-cpp-release-b9693-adds-bf16-support-and-cross-platform-binaries/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1723-llama-cpp-release-b9693-adds-bf16-support-and-cross-platform-binaries/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1725-crewai-v1-14-8a-vypustil-novye-funktsii-flowdefinition/2026-06-18weekly0.7 https://korshunov.ai/en/article/1725-crewai-v1-14-8a-releases-new-flowdefinition-features/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1724-u-kogo-est-dostatochno-vychislitelnykh-resursov-chtoby-sozdat-distilliatsionnyi/2026-06-18weekly0.7 https://korshunov.ai/en/article/1722-llama-cpp-releases-version-b9694-with-new-binaries-for-multiple-platforms/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1722-llama-cpp-vypustil-versiiu-b9694-s-novymi-binarnymi-failami-dlia-neskolkikh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1718-localllama-proposes-crowdsourced-coding-dataset/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1718-localllama-predlagaet-dataset-dlia-programmirovaniia-sozdannyi-soobshchestvom/2026-06-18weekly0.7 https://korshunov.ai/en/article/1715-glm-5-2-review-and-censorship-response/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1716-chto-vy-nedavno-zanimalis/2026-06-18weekly0.7 https://korshunov.ai/en/article/1716-what-have-you-been-working-on-lately/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1715-obzor-glm-5-2-i-reaktsiia-na-tsenzuru/2026-06-18weekly0.7 https://korshunov.ai/en/article/1708-price-rising-effect-is-wild/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1709-midjourney-predstavliaet-skaner-ultrazvukovoi-tomografii-tela/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1708-effekt-rosta-tseny-iavliaetsia-neveroiatnym/2026-06-18weekly0.7 https://korshunov.ai/en/article/1709-midjourney-launches-full-body-ultrasound-ct-scanner/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1706-diskriminator-orientirovannyi-rl-korrektiruet-potokovoe-sootvetstvie-s-nagrad-s/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1705-sliianie-osnovnykh-podprostranstv-dlia-mnogozadachnogo-obucheniia/2026-06-18weekly0.7 https://korshunov.ai/en/article/1706-discriminator-guided-rl-corrects-flow-matching-with-data-aligned-rewards/2026-06-18weekly0.7 https://korshunov.ai/en/article/1705-essential-subspace-merging-for-multi-task-learning/2026-06-18weekly0.7 https://korshunov.ai/en/article/1703-batch-size-tradeoffs-in-stochastic-momentum-methods/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1703-protivorechiia-po-razmeru-batchei-v-stokhasticheskikh-metodakh-impulsa/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1701-agdn-reshenie-zadachi-o-prodavtse-puteshestvii-s-anizotropnym-grafovym-metodom/2026-06-18weekly0.7 https://korshunov.ai/en/article/1701-agdn-solving-traveling-salesman-problem-with-anisotropic-graph-diffusion/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1704-bezopasnost-otrazheniia-v-predobuchenii-dlia-iazykovykh-modelei/2026-06-18weekly0.7 https://korshunov.ai/en/article/1704-safety-reflection-pretraining-for-llms/2026-06-18weekly0.7 https://korshunov.ai/en/article/1702-cross-dataset-auc-for-realistic-deepfake-detector-evaluation/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1700-avtomatizirovannyi-freimvork-dlia-zaderzhannykh-i-lozhnykh-triggerov-sistemy/2026-06-18weekly0.7 https://korshunov.ai/en/article/1700-automated-annotation-framework-for-delayed-and-false-aeb-triggers/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1702-perekrestnaia-otsenka-auc-dlia-otsenki-realistichnogo-detektora-feikov/2026-06-18weekly0.7 https://korshunov.ai/en/article/1698-generalised-eigenvalue-geometry-of-semantic-adversarial-attacks/2026-06-18weekly0.7 https://korshunov.ai/en/article/1697-xgboost-forget-for-machine-unlearning-in-network-intrusion-detection/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1699-reshenie-orientirovannyi-rl-dlia-zariadki-elektromobilei-s-neizvestnymi-vyezda/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1697-xgboost-forget-dlia-mashinnogo-zabveniia-v-obnaruzhenii-setevykh-intsidentov/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1698-obshchaia-geometriia-sobstvennykh-znachenii-semanticheskikh-atak/2026-06-18weekly0.7 https://korshunov.ai/en/article/1699-decision-focused-rl-for-ev-charging-with-unknown-departure-times/2026-06-18weekly0.7 https://korshunov.ai/en/article/1696-mast-enables-selective-unlearning-in-rlvr-induced-reasoning/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1696-mast-obespechivaet-selektivnoe-iskliuchenie-v-protsesse-priniatiia-reshenii-rlvr/2026-06-18weekly0.7 https://korshunov.ai/en/article/1694-stare-surprisal-guided-token-level-advantage-reweighting-for-policy-entropy/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1694-stare-regulirovanie-preimushchestv-na-urovne-tokenov-s-ispolzovaniem-siurpriza/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1695-baiesovskaia-optimizatsiia-s-uchastiem-cheloveka-dlia-ogranichenii-bioprotsessov/2026-06-18weekly0.7 https://korshunov.ai/en/article/1695-bayesian-optimization-with-human-in-the-loop-for-bioprocess-constraints/2026-06-18weekly0.7 https://korshunov.ai/en/article/1692-tgo-i-spectral-geometry-of-vision-transformers/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1693-txbench-pp-proizvoditelnost-ii-agenta-v-presledovanii-farmakologii/2026-06-18weekly0.7 https://korshunov.ai/en/article/1691-graph-neural-networks-accelerate-algebraic-multigrid-pressure-solver/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1692-tgo-i-spektralnaia-geometriia-vizualnykh-transformerov/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1691-grafovye-neironnye-seti-uskoriaiut-reshateli-na-osnove-algebraicheskoi-multigrid/2026-06-18weekly0.7 https://korshunov.ai/en/article/1693-txbench-pp-ai-agent-performance-in-preclinical-pharmacology/2026-06-18weekly0.7 https://korshunov.ai/en/article/1686-conceptual-innovation-in-medical-imaging-ai/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1690-onecanvas-ponimanie-3d-stseny-cherez-pereproektsiiu-panoramy/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1686-kontseptualnaia-innovatsiia-v-iskusstvennom-intellekte-meditsinskoi/2026-06-18weekly0.7 https://korshunov.ai/en/article/1689-scan-multi-scale-clustering-for-time-series-anomaly-detection/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1689-scan-mnogourovnevoe-klasternoe-modelirovanie-dlia-obnaruzheniia-anomalii-v/2026-06-18weekly0.7 https://korshunov.ai/en/article/1690-onecanvas-3d-scene-understanding-via-panoramic-reprojection/2026-06-18weekly0.7 https://korshunov.ai/en/article/1687-large-language-gibbs-for-structured-probabilistic-inference/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1688-nulevoe-vliianie-na-monitoring-vyiavliaet-skrytye-trenirovki-mashinnogo/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1687-bolshie-iazykovye-gibbs-dlia-strukturirovannoi-veroiatnostnoi-inferentsii/2026-06-18weekly0.7 https://korshunov.ai/en/article/1688-zero-overhead-telemetry-detects-hidden-ml-training/2026-06-18weekly0.7 https://korshunov.ai/en/article/1684-ambient-sound-and-light-predict-icu-delirium/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1685-nesycat-torch-realizatsiia-differentsiruemykh-tenzorov-dlia-neirosimvolnogo/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1684-zvuk-i-svet-predskazyvaiut-delirium-v-itsu/2026-06-18weekly0.7 https://korshunov.ai/en/article/1685-nesycat-torch-differentiable-tensor-implementation-for-neurosymbolic-learning/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1683-act2answer-otsenivaet-sokhranenie-znanii-v-modeliakh-vizualno-iazykovykh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1683-act2answer-evaluates-knowledge-retention-in-vision-language-action-models/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1680-p-k-gcn-fizicheski-usilennyi-koopman-usilennyi-grafovyi-konvoliutsionnyi-setei/2026-06-18weekly0.7 https://korshunov.ai/en/article/1680-p-k-gcn-physics-augmented-koopman-enhanced-graph-convolutional-network/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1681-optimizatsiia-klimaticheskikh-stsenariev-povyshaet-obobshchenie-emuliatora/2026-06-18weekly0.7 https://korshunov.ai/en/article/1682-mc-dropout-uncertainty-alignment-insufficient-for-clinical-safety-in-glioma/2026-06-18weekly0.7 https://korshunov.ai/en/article/1681-optimizing-climate-scenarios-boosts-emulator-generalization/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1682-nedostatochnost-sinkhronizatsii-neopredelennosti-mc-dropout-dlia-klinicheskoi-v/2026-06-18weekly0.7 https://korshunov.ai/en/article/1679-diffusion-proof-first-framework-for-diffusion-llms-in-formal-theorem-proving/2026-06-18weekly0.7 https://korshunov.ai/en/article/1678-reverse-engineering-transformer-attention-with-executable-programs/2026-06-18weekly0.7 https://korshunov.ai/en/article/1677-ubp2-uncertainty-balanced-preference-planning-for-efficient-preference-based-rl/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1677-ubp2-metod-sbalansirovannogo-ucheta-neopredelennosti-dlia-effektivnogo-s/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1679-diffusion-proof-pervyi-freimvork-dlia-diffuzionnykh-llm-v-formalnoi-dokazatelnoi/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1678-razbor-vnimaniia-transformera-s-pomoshchiu-ispolniaemykh-programm/2026-06-18weekly0.7 https://korshunov.ai/en/article/1675-locus-a-local-ordinance-corpus-for-the-united-states/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1674-lidery-anthropic-i-google-deepmind-prizyvaiut-k-sozdaniiu-koalitsii-po-ii-pod-na/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1675-locus-lokalnaia-kollektsiia-zakonov-dlia-ssha/2026-06-18weekly0.7 https://korshunov.ai/en/article/1676-chandra-gaia-catalog-uses-machine-learning-to-resolve-x-ray-and-optical-source/2026-06-18weekly0.7 https://korshunov.ai/en/article/1674-anthropic-and-google-deepmind-ceos-call-for-u-s-led-ai-coalition-at-g7/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1676-katalog-chandra-gaia-ispolzuet-mashinnoe-obuchenie-dlia-razresheniia-mezhdu-i/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1672-bezopasnost-otrazheniia-v-predobuchenii-dlia-iazykovykh-modelei/2026-06-18weekly0.7 https://korshunov.ai/en/article/1671-user-as-engram-local-parametric-edits-for-personal-memory/2026-06-18weekly0.7 https://korshunov.ai/en/article/1673-essential-subspace-merging-for-multi-task-learning/2026-06-18weekly0.7 https://korshunov.ai/en/article/1672-safety-reflection-pretraining-for-llms/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1673-sliianie-osnovnykh-podprostranstv-dlia-mnogozadachnogo-obucheniia/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1671-polzovatel-kak-engra-lokalnye-parametricheskie-redaktsii-dlia-lichnoi-pamiati/2026-06-18weekly0.7 https://korshunov.ai/en/article/1670-clinician-centered-pipeline-for-ultrasound-ai-annotation-and-evaluation/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1669-operatsionno-validirovannaia-sistema-vizualnogo-tsikla-dlia-avtonomnosti-dronov/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1668-protivorechiia-po-razmeru-batchei-v-stokhasticheskikh-metodakh-impulsa/2026-06-18weekly0.7 https://korshunov.ai/en/article/1668-batch-size-tradeoffs-in-stochastic-momentum-methods/2026-06-18weekly0.7 https://korshunov.ai/en/article/1669-hardware-validated-vision-in-the-loop-for-maritime-uav-autonomy/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1670-potok-orientirovannyi-na-klinitsistov-dlia-annotatsii-i-otsenki-iskusstvennogo-v/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1665-rl-s-fokusom-na-resheniia-dlia-zariadki-elektromobilei-s-neizvestnymi-vremenami/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1667-llm-kak-interfeis-ml-kak-prognozator-dlia-detskoi-appenditsita/2026-06-18weekly0.7 https://korshunov.ai/en/article/1667-llm-as-interface-ml-as-predictor-for-pediatric-appendicitis/2026-06-18weekly0.7 https://korshunov.ai/en/article/1666-combining-properties-in-elbot-abox-abduction/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1666-sochetanie-svoistv-v-elbot-abox-abduktsii/2026-06-18weekly0.7 https://korshunov.ai/en/article/1665-decision-focused-rl-for-ev-charging-with-unknown-departure-times/2026-06-18weekly0.7 https://korshunov.ai/en/article/1664-xgboost-forget-for-machine-unlearning-in-network-intrusion-detection/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1664-xgboost-forget-dlia-mashinnogo-zabveniia-v-sistemakh-obnaruzheniia-setevykh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1662-stare-surprisal-guided-token-level-advantage-reweighting-for-policy-entropy/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1662-stare-regulirovanie-preimushchestv-na-urovne-tokenov-po-printsipu-neozhidannosti/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1663-mast-obespechivaet-selektivnoe-ischeznovenie-v-protsesse-priniatiia-reshenii/2026-06-18weekly0.7 https://korshunov.ai/en/article/1663-mast-enables-selective-unlearning-in-rlvr-induced-reasoning/2026-06-18weekly0.7 https://korshunov.ai/en/article/1661-txbench-pp-ai-agent-benchmark-in-preclinical-pharmacology/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1660-klassifikatsiia-sviazyvaet-potrebnosti-prismatrivaiushchikh-s-tekhnologiiami-v/2026-06-18weekly0.7 https://korshunov.ai/en/article/1660-taxonomy-links-caregiver-needs-to-mental-health-tech/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1657-mnogodomennyi-benchmark-dlia-obnaruzheniia-tekstovykh-izobrazhenii-ii/2026-06-18weekly0.7 https://korshunov.ai/en/article/1657-multi-domain-benchmark-for-detecting-ai-generated-text-rich-images/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1661-txbench-pp-otsenka-ai-agenta-v-proklinicheskoi-farmakologii/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1658-x-slides-metrika-dlia-generatsii-slaidov-s-uchetom-potrebnostei-auditorii/2026-06-18weekly0.7 https://korshunov.ai/en/article/1658-x-slides-benchmark-for-audience-conditioned-slide-generation/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1659-onecanvas-ponimanie-3d-stseny-cherez-pereproektsiiu-panoramy/2026-06-18weekly0.7 https://korshunov.ai/en/article/1659-onecanvas-3d-scene-understanding-via-panoramic-reprojection/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1655-nesycat-torch-realizatsiia-differentsiruemykh-tenzorov-dlia-neirosimvolnogo/2026-06-18weekly0.7 https://korshunov.ai/en/article/1656-trade-offs-in-medical-llm-adaptation-french-qa-study/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1654-samokorrektsiia-povyshaet-doverie-k-sotsialnym-chat-botam/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1656-protivorechiia-v-adaptatsii-meditsinskikh-llm-issledovanie-po-voprosam-v-oblasti/2026-06-18weekly0.7 https://korshunov.ai/en/article/1654-self-correction-boosts-trust-in-social-chatbots/2026-06-18weekly0.7 https://korshunov.ai/en/article/1655-nesycat-torch-differentiable-tensor-implementation-for-neurosymbolic-learning/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1652-agenty-intellektualnogo-analiza-dannykh-obespechivaiut-avtonomnyi-zapros-dannykh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1652-data-intelligence-agents-enable-autonomous-data-querying/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1651-scena-sistema-generatsii-audio-stseny-na-osnove-ssylok/2026-06-18weekly0.7 https://korshunov.ai/en/article/1651-scena-reference-driven-multi-speaker-audio-scene-generation/2026-06-18weekly0.7 https://korshunov.ai/en/article/1653-reverse-engineering-transformer-attention-with-executable-programs/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1653-razbor-vnimaniia-transformera-s-pomoshchiu-ispolniaemykh-programm/2026-06-18weekly0.7 https://korshunov.ai/en/article/1650-rubric-conditioned-self-distillation-framework/2026-06-18weekly0.7 https://korshunov.ai/en/article/1649-ubp2-uncertainty-balanced-preference-planning-for-efficient-preference-based-rl/2026-06-18weekly0.7 https://korshunov.ai/en/article/1648-benchmarking-small-llms-on-hard-html-data-extraction/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1649-ubp2-metod-sbalansirovannogo-ucheta-neopredelennosti-dlia-effektivnogo-s/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1650-kadrovo-uslovnaia-samodistilliatsiia/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1648-otsenka-malykh-modelei-llm-na-slozhnykh-dannykh-iz-html/2026-06-18weekly0.7 https://korshunov.ai/en/article/1645-dango-a-strictly-l1-only-llm-for-sla-research/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1645-dango-strogo-odnolingvalnyi-llm-dlia-issledovanii-sla/2026-06-18weekly0.7 https://korshunov.ai/en/article/1644-llm-as-interface-ml-as-predictor-for-pediatric-appendicitis/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1644-llm-kak-interfeis-ml-kak-prognozator-dlia-detskoi-appenditsitnoi-bolezni/2026-06-18weekly0.7 https://korshunov.ai/en/article/1643-recom-validity-discrimination-tradeoff-in-reddit-qa-metrics/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1643-recom-sviaz-mezhdu-dostovernostiu-i-diskriminatsiei-v-metrikakh-dlia-zadachi-na/2026-06-18weekly0.7 https://korshunov.ai/en/article/1640-large-language-gibbs-for-structured-probabilistic-inference/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1641-dreamreasoner-8b-obuchenie-kurrikulumom-po-razmeru-blokov-dlia-rassuzhdeniia-na/2026-06-18weekly0.7 https://korshunov.ai/en/article/1641-dreamreasoner-8b-block-size-curriculum-learning-for-diffusion-reasoning/2026-06-18weekly0.7 https://korshunov.ai/en/article/1642-stare-surprisal-guided-token-level-advantage-reweighting-for-policy-entropy/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1642-stare-surprisal-guided-token-level-advantage-reweighting-for-policy-entropy/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1640-bolshie-iazykovye-gibbs-dlia-strukturirovannogo-veroiatnostnogo-vyvoda/2026-06-18weekly0.7 https://korshunov.ai/en/article/1639-empirical-study-of-medical-llm-adaptation-in-french-qa/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1638-mnogokanalnyi-fiktivnyi-khod-dlia-priniatiia-reshenii-s-vovlecheniem-pozitsii/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1637-kadrovo-uslovnaia-samodistilliatsiia/2026-06-18weekly0.7 https://korshunov.ai/en/article/1638-multi-agent-fictitious-play-for-stance-entangled-decision-making/2026-06-18weekly0.7 https://korshunov.ai/en/article/1637-rubric-conditioned-self-distillation-framework/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1639-empiricheskoe-issledovanie-adaptatsii-meditsinskikh-llm-na-frantsuzskom-iazyke/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1636-locus-lokalnaia-korporatsiia-normativnykh-aktov-dlia-soedinennykh-shtatov/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1635-turing-rl-obuchenie-imitatorov-polzovatelei-s-voznagrazhdeniiami-tiuringa/2026-06-18weekly0.7 https://korshunov.ai/en/article/1635-turing-rl-learning-user-simulators-with-turing-rewards/2026-06-18weekly0.7 https://korshunov.ai/en/article/1636-locus-a-local-ordinance-corpus-for-the-united-states/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1634-omniagent-integrirovannaia-aktivnaia-vospriiatie-dlia-mnogomodalnogo-ponimaniia/2026-06-18weekly0.7 https://korshunov.ai/en/article/1634-omniagent-native-active-perception-for-omni-modal-understanding/2026-06-18weekly0.7 https://korshunov.ai/en/article/1631-detecting-structural-biases-via-causal-mechanism-shifts/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1631-obnaruzhenie-strukturnykh-iskazhenii-cherez-sdvigi-prichinnykh-mekhanizmov/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1632-semi-nabliudaemoe-klasternoe-modelirovanie-s-ispolzovaniem-a-kontra-anomalii/2026-06-18weekly0.7 https://korshunov.ai/en/article/1632-seed-guided-semi-supervised-clustering-via-a-contrario-anomaly-detection/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1630-skill-mas-evoliutsionnaia-metanavyk-dlia-avtomaticheskikh-sistem-mnogoagentov/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1628-tapo-samodistilliatsiia-s-mikrosootvetstvuiushchimi-traektoriiami/2026-06-18weekly0.7 https://korshunov.ai/en/article/1629-semantic-robustness-certification-for-vision-language-models/2026-06-18weekly0.7 https://korshunov.ai/en/article/1628-tapo-self-distillation-with-micro-reflective-trajectories/2026-06-18weekly0.7 https://korshunov.ai/en/article/1630-skill-mas-evolving-meta-skill-for-automatic-multi-agent-systems/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1629-semanticheskaia-ustoichivost-sertifikatsii-dlia-modelei-vizualno-iazykovykh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1625-inductive-biases-in-ml-emulation-of-sudden-stratospheric-warmings/2026-06-18weekly0.7 https://korshunov.ai/en/article/1627-kernel-of-partition-paths-unified-representation-for-tree-ensembles/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1625-induktivnye-predvziatosti-v-mashinnom-obuchenii-pri-emuliatsii-rezkikh/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1627-kernel-putei-razdeleniia-edinaia-predstavlenie-dlia-derevev-ansamblei/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1626-priblizhennaia-strukturirovannaia-diffuziia-dlia-posledovatelnoi-annotatsii/2026-06-18weekly0.7 https://korshunov.ai/en/article/1626-approximate-structured-diffusion-for-sequence-labelling/2026-06-18weekly0.7 https://korshunov.ai/en/article/1622-feature-selection-and-ridge-regularization-in-strategic-classification/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1622-vybor-priznakov-i-reguliarizatsiia-ridzha-v-strategicheskoi-klassifikatsii/2026-06-18weekly0.7 https://korshunov.ai/en/article/1624-quam-sm-framework-for-uncertainty-quantification-in-medical-image-segmentation/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1624-kadr-sm-freimvork-dlia-otsenki-neopredelennosti-v-meditsinskoi-segmentatsii/2026-06-18weekly0.7 https://korshunov.ai/en/article/1623-scaling-aeb-with-unlabeled-data-via-meta-feedback-ssl/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1623-rasshirenie-sistemy-avariinogo-tormozheniia-s-ispolzovaniem-neprimeniaemykh-meta/2026-06-18weekly0.7 https://korshunov.ai/en/article/1619-latent-sdes-for-anomaly-detection-in-sparse-multivariate-time-series/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1620-neironnye-seti-s-uchetom-sdviga-domena-dlia-otsenki-nesbalansirovannoi-massy/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1621-proverka-adaptatsii-na-etape-testirovaniia-s-ispolzovaniem-metoda-potokovogo-v/2026-06-18weekly0.7 https://korshunov.ai/en/article/1621-flow-matching-test-time-adaptation-for-oct-image-denoising/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1619-latentnye-sde-dlia-obnaruzheniia-anomalii-v-redkikh-mnogomernykh-vremennykh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1620-domain-shift-aware-neural-networks-for-unbalance-mass-estimation/2026-06-18weekly0.7 https://korshunov.ai/en/article/1618-reves-augmented-training-for-test-time-scaling/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1616-grapnet-programmiruemaia-dinamicheskaia-arkhitektura-neironnoi-grafy/2026-06-18weekly0.7 https://korshunov.ai/en/article/1616-grapnet-a-programmable-dynamic-architecture-neural-graph-substrate/2026-06-18weekly0.7 https://korshunov.ai/en/article/1617-complexity-results-for-binarized-neural-network-robustness-verification/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1617-rezultaty-slozhnosti-dlia-ustoichivosti-proverki-binarizirovannykh-neironnykh/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1618-reves-usilennoe-obuchenie-dlia-masshtabirovaniia-na-etape-testirovaniia/2026-06-18weekly0.7 https://korshunov.ai/en/article/1613-unsupervised-reward-optimization-for-protein-language-models/2026-06-18weekly0.7 https://korshunov.ai/en/article/1614-zero-shot-active-feature-acquisition-via-llm-elicitation/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1613-nepodgotovlennaia-optimizatsiia-voznagrazhdeniia-dlia-iazykovykh-modelei-belkov/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1615-transitnet-dostigaet-tochnosti-952-pri-poiske-tranzitov-v-usloviiakh-nizkogo/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1614-nulevaia-aktivnaia-akusticheskaia-poluchenie-priznakov-cherez-elicitatsiiu-llm/2026-06-18weekly0.7 https://korshunov.ai/en/article/1615-transitnet-achieves-95-2-accuracy-in-low-snr-transit-searches/2026-06-18weekly0.7 https://korshunov.ai/en/article/1602-efficientrollout-system-aware-self-speculative-decoding-for-rl-rollouts/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1607-obuchenie-bez-voznagrazhdeniia-na-vizualnykh-potokakh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1607-reward-free-learning-from-perceptual-streams/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1602-efficientrollout-sistemno-osoznannaia-samospekuliativnaia-dekodirovka-dlia-rl/2026-06-18weekly0.7 https://korshunov.ai/en/article/1600-quantum-gan-augmentation-shows-no-benefit-in-brain-mri/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1600-kvantovoe-rasshirenie-gan-ne-pokazyvaet-preimushchestv-v-mrt-mozga/2026-06-18weekly0.7 https://korshunov.ai/en/article/1596-fosc-x-framework-for-top-m-optimal-flat-clusterings/2026-06-18weekly0.7 https://korshunov.ai/en/article/1594-robust-sequential-conditional-independence-testing/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1596-fosc-x-freimvork-dlia-optimalnykh-ploskikh-klasterizatsii-na-urovne-m/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1594-ustoichivyi-posledovatelnyi-test-uslovnoi-nezavisimosti/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1585-v-fokuse-ispolzovanie-gpu-spot-dlia-uskoreniia-postobuchivaniia-dit-rl/2026-06-18weekly0.7 https://korshunov.ai/en/article/1582-sumi-open-uniform-diffusion-language-model-from-scratch/2026-06-18weekly0.7 https://korshunov.ai/en/article/1592-diphine-neural-estimator-for-ph-id-in-continuous-systems/2026-06-18weekly0.7 https://korshunov.ai/en/article/1585-spotlight-using-spot-gpus-to-accelerate-dit-rl-post-training/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1582-sumi-otkrytaia-unifitsirovannaia-model-raspredelennoi-generatsii-iazyka-s-nulia/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1592-diphine-neironnyi-otsenshchik-dlia-ph-id-v-nepreryvnykh-sistemakh/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1579-moat-dinamicheskii-analiz-s-uchetom-zhiznennogo-tsikla-dlia-bezopasnoi-modelei/2026-06-18weekly0.7 https://korshunov.ai/en/article/1572-lstm-vision-transformer-improves-hrrr-forecast-error-prediction/2026-06-18weekly0.7 https://korshunov.ai/en/article/1576-fomoe-breaks-full-replica-barrier-with-partitioned-expert-layers/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1576-fomoe-preodolevaet-porog-polnykh-kopii-s-pomoshchiu-razdelennykh-ekspertnykh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1579-moat-lifecycle-aware-dynamic-analysis-for-secure-ml-model-execution/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1572-lstm-videnie-transformer-uluchshaet-predskazanie-oshibok-prognoza-hrrr/2026-06-18weekly0.7 https://korshunov.ai/en/article/1566-positive-unlabeled-learning-for-llm-evaluation-auditing/2026-06-18weekly0.7 https://korshunov.ai/en/article/1567-learnable-speech-to-spike-encoder-for-spiking-neural-networks/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1568-geometricheskii-i-stokhasticheskii-analiz-razryvov-v-razrezhennykh-modeliakh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1568-geometric-and-stochastic-analysis-of-discontinuities-in-sparse-mixture-of/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1566-obuchenie-s-polozhitelnymi-i-neprimechaemymi-primerami-dlia-audita-otsenki/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1567-obuchaemyi-kodirovshchik-rechi-spaikov-dlia-spaikovykh-neironnykh-setei/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1565-tipy-voznagrazhdenii-v-rl-povyshaiut-ustoichivost-v-kiber-fizicheskikh-sistemakh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1565-rl-reward-types-enhance-resilience-in-cyber-physical-systems/2026-06-18weekly0.7 https://korshunov.ai/en/article/1553-context-aware-follow-up-optimization-for-type-2-diabetes/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1553-optimizatsiia-intervalov-posleduiushchikh-vizitov-s-uchetom-konteksta-dlia-2/2026-06-18weekly0.7 https://korshunov.ai/en/article/1551-structure-first-architectures-for-dynamical-learning/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1551-struktura-pervye-arkhitektury-dlia-dinamicheskogo-obucheniia/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1545-wasserstein-policy-learning-for-distributional-outcomes/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1549-plavnost-orientirovannaia-dereandomizatsiia-granits-pac-bayes/2026-06-18weekly0.7 https://korshunov.ai/en/article/1548-journeyformer-sequence-modeling-for-airbnb-guest-journeys/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1548-journeyformer-modelirovanie-posledovatelnostei-dlia-puteshestvii-gostei-airbnb/2026-06-18weekly0.7 https://korshunov.ai/en/article/1545-wasserstein-policy-learning-for-distributional-outcomes/2026-06-18weekly0.7 https://korshunov.ai/en/article/1549-smoothness-based-derandomization-of-pac-bayes-bounds/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1544-xai-raskryvaet-kliuchevye-faktory-v-evropeiskikh-elektricheskikh-rynkakh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1544-xai-reveals-key-drivers-in-european-electricity-markets/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1543-vigos-razdelenie-vospriiatiia-i-rassuzhdeniia-v-mnogomodalnykh-na-politicheskikh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1543-vigos-decoupling-perception-and-reasoning-in-multimodal-on-policy-self/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1540-pareto-q-learning-s-mashinami-voznagrazhdenii/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1542-giskard-protokol-agregatsii-s-konfidentsialnostiu-i-ustoichivostiu-k-uzlam/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1539-indeqs-graf-orientirovannye-neironnye-uravneniia-differentsialnykh-uravnenii/2026-06-18weekly0.7 https://korshunov.ai/en/article/1542-giskard-confidential-and-byzantine-robust-aggregation-protocol/2026-06-18weekly0.7 https://korshunov.ai/en/article/1540-pareto-q-learning-with-reward-machines/2026-06-18weekly0.7 https://korshunov.ai/en/article/1539-indeqs-graph-informed-neural-controlled-differential-equations/2026-06-18weekly0.7 https://korshunov.ai/en/article/1535-local-population-risk-certificates-for-model-updates/2026-06-18weekly0.7 https://korshunov.ai/en/article/1537-chronosurv-a-graph-framework-for-multimodal-survival-analysis/2026-06-18weekly0.7 https://korshunov.ai/en/article/1536-orthoreg-orthogonal-regularization-for-hybrid-symbolic-neural-dynamical-systems/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1537-chronosurv-freimvork-grafa-dlia-analiza-vyzhivaemosti-v-mnogomodalnykh-dannykh/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1535-lokalnye-sertifikaty-riska-dlia-obnovleniia-modelei/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1536-orthoreg-ortogonalnaia-reguliarizatsiia-dlia-gibridnykh-simvolicheskikh-sistem/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1533-cahp-komplementarnoe-uproshchenie-golov-vnimaniia-dlia-effektivnykh/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1612-chestnoe-obnaruzhenie-kognitivnykh-narushenii-cherez-otuchivanie/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1534-openant-sistema-obnaruzheniia-uiazvimostei-na-osnove-llm/2026-06-18weekly0.7 https://korshunov.ai/en/article/1612-fair-cognitive-impairment-detection-through-unlearning/2026-06-18weekly0.7 https://korshunov.ai/en/article/1533-cahp-complementary-attention-head-pruning-for-efficient-transformers/2026-06-18weekly0.7 https://korshunov.ai/en/article/1534-openant-llm-powered-vulnerability-discovery-system/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1611-rech-obuslovlennaia-konechnaia-sistema-raspoznavaniia-iazykov-dlia-kitaiskikh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1611-speech-driven-end-to-end-language-discrimination-for-chinese-dialects/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1610-rasstoianie-adaptivnoe-predstavlenie-dlia-vnimaniia/2026-06-18weekly0.7 https://korshunov.ai/en/article/1609-cddtlda-transfer-learning-for-chinese-dialect-discrimination/2026-06-18weekly0.7 https://korshunov.ai/en/article/1610-distance-adaptive-representation-for-attention/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1609-cddtlda-perenos-obucheniia-dlia-diskriminatsii-kitaiskikh-dialektov/2026-06-18weekly0.7 https://korshunov.ai/en/article/1606-physassistbench-evaluates-llms-in-doctor-patient-ehr-interaction/2026-06-18weekly0.7 https://korshunov.ai/en/article/1605-bcl-bayesian-in-context-learning-for-information-extraction/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1606-physassistbench-otsenivaet-llms-v-vzaimodeistvii-doktor-patsient-ekhr/2026-06-18weekly0.7 https://korshunov.ai/en/article/1608-steerable-cultural-preference-optimization-of-reward-models/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1608-reguliruemaia-optimizatsiia-kulturnykh-predpochtenii-v-modeliakh/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1605-bcl-baiesovskoe-obuchenie-v-kontekste-dlia-izvlecheniia-informatsii/2026-06-18weekly0.7 https://korshunov.ai/en/article/1601-misfired-alignment-in-llms-a-quantitative-study/2026-06-18weekly0.7 https://korshunov.ai/en/article/1603-pec-home-simulated-dataset-for-elliptical-command-interpretation/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1603-pec-home-simulirovannyi-dataset-dlia-interpretatsii-ellipticheskikh-komand/2026-06-18weekly0.7 https://korshunov.ai/en/article/1604-pragrest-self-reinforcing-counterfactual-reasoning-for-pragmatic-language/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1604-pragrest-samorazvivaiushchaiasia-kosvennaia-logika-dlia-ponimaniia-priamogo/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1601-nepravilnaia-sinkhronizatsiia-v-bolshikh-iazykovykh-modeliakh-kolichestvennoe/2026-06-18weekly0.7 https://korshunov.ai/en/article/1599-regmix-d-dynamic-data-mixing-via-proxy-training-trajectories/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1599-regmix-d-dinamicheskoe-smeshivanie-dannykh-za-schet-traektorii-obucheniia-cherez/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1597-forecastbench-sim-benchmark-prognozirovaniia-v-imitatsionnom-mire/2026-06-18weekly0.7 https://korshunov.ai/en/article/1597-forecastbench-sim-simulated-world-forecasting-benchmark/2026-06-18weekly0.7 https://korshunov.ai/en/article/1598-ears-framework-enhances-multi-agent-system-reliability/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1598-kadr-ears-povyshaet-nadezhnost-mnogoagentnykh-sistem/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1593-tw-legalbench-otsenka-llm-na-taivanskom-prave/2026-06-18weekly0.7 https://korshunov.ai/en/article/1593-tw-legalbench-evaluating-llms-on-taiwanese-law/2026-06-18weekly0.7 https://korshunov.ai/en/article/1591-llms-struggle-to-capture-item-discrimination-in-reading-assessments/2026-06-18weekly0.7 https://korshunov.ai/en/article/1595-frustrated-synchronization-network-outperforms-transformers/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1595-seti-s-razdrazhennoi-sinkhronizatsiei-prevoskhodiat-transformery/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1591-llms-ne-spravliaiutsia-s-vyiavleniem-razlichii-mezhdu-elementami-v-otsenkakh/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1590-morpheus-neironnyi-tokenizator-i-vektorizator-dlia-turetskogo-iazyka/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1588-redaktirovanie-vektora-vykhoda-snizhaet-zapominanie-v-iazykovykh-modeliakh/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1589-legalworld-tsiklicheskaia-sreda-dlia-iuridicheskikh-agentov/2026-06-18weekly0.7 https://korshunov.ai/en/article/1589-legalworld-life-cycle-environment-for-legal-agents/2026-06-18weekly0.7 https://korshunov.ai/en/article/1588-output-vector-editing-reduces-memorization-in-llms/2026-06-18weekly0.7 https://korshunov.ai/en/article/1590-morpheus-neural-tokenizer-and-embedder-for-turkish/2026-06-18weekly0.7 https://korshunov.ai/en/article/1586-dice-improves-long-document-retrieval-with-chunk-evidence-aggregation/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1587-sama-edinaia-platforma-dlia-rasshireniia-multimodalnykh-dannykh-v-usloviiakh/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1586-dice-uluchshaet-poisk-v-dlinnykh-dokumentakh-s-agregatsiei-dokazatelstv-po/2026-06-18weekly0.7 https://korshunov.ai/en/article/1587-sama-unified-framework-for-low-resource-multimodal-data-augmentation/2026-06-18weekly0.7 https://korshunov.ai/en/article/1584-redactionbench-a-benchmark-for-contextual-privacy-in-ai/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1583-handwritingagent-sintez-ruchnogo-pocherka-na-osnove-iazyka-v-svg/2026-06-18weekly0.7 https://korshunov.ai/en/article/1583-handwritingagent-language-driven-handwriting-synthesis-in-svg/2026-06-18weekly0.7 https://korshunov.ai/en/article/1581-llm-based-metrics-improve-clinical-significance-evaluation-in-radiology/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1581-metriki-na-osnove-bolshikh-iazykovykh-modelei-uluchshaiut-otsenku-klinicheskoi-v/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1584-redactionbench-benchmark-dlia-kontekstualnoi-privatnosti-v-ii/2026-06-18weekly0.7 https://korshunov.ai/en/article/1578-data-recipe-boosts-long-context-reasoning-in-llms/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1577-scholarsum-summarizatsiia-student-uchitel-cherez-logicheskoe-obosnovanie-znanii/2026-06-18weekly0.7 https://korshunov.ai/en/article/1577-scholarsum-student-teacher-abstractive-summarization-via-knowledge-graph/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1580-gatemem-otsenka-upravleniia-pamiatiu-v-agentakh-s-obshchei-pamiatiu-v-interesakh/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1578-dannye-retseptura-povyshaet-dolgosrochnoe-myshlenie-v-bolshikh-iazykovykh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1580-gatemem-benchmarking-memory-governance-in-multi-principal-shared-memory-agents/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1575-impsh-uluchshaet-obnaruzhenie-skrytogo-religioznogo-spektra-v-raznykh-oblastiakh/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1573-distilliatsiia-s-sinteticheskimi-dannymi-dlia-analiza-finansovogo-nastroeniia/2026-06-18weekly0.7 https://korshunov.ai/en/article/1573-distillation-with-synthetic-data-for-financial-sentiment-analysis/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1574-priblizhennaia-strukturirovannaia-diffuziia-dlia-posledovatelnoi-metki/2026-06-18weekly0.7 https://korshunov.ai/en/article/1574-approximate-structured-diffusion-for-sequence-labelling/2026-06-18weekly0.7 https://korshunov.ai/en/article/1575-impsh-improves-implicit-hate-speech-detection-across-domains/2026-06-18weekly0.7 https://korshunov.ai/en/article/1569-sage-stochastic-prompt-optimization-via-agent-guided-exploration/2026-06-18weekly0.7 https://korshunov.ai/en/article/1571-rubric-guided-counterfactual-recommendations-for-medical-communication/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1570-rpcl-uluchshaet-izvlechenie-par-emotsii-prichin/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1571-rubrik-orientirovannye-kontrfaktnye-rekomendatsii-dlia-meditsinskoi/2026-06-18weekly0.7 https://korshunov.ai/en/article/1570-rpcl-improves-multimodal-emotion-cause-pair-extraction/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1569-sage-sluchainaia-optimizatsiia-promptov-s-pomoshchiu-eksperimentalnogo-agentov/2026-06-18weekly0.7 https://korshunov.ai/en/article/1562-graph-esbmc-plc-formal-verification-of-graphical-plcopen-ld-programs/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1562-graph-esbmc-plc-formalnaia-proverka-graficheskikh-programm-plcopen-ld/2026-06-18weekly0.7 https://korshunov.ai/en/article/1564-reves-augmented-training-for-test-time-scaling/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1563-llms-stalkivaiutsia-s-otritsaniem-v-metaforicheskoi-rechi/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1564-reves-usilennoe-obuchenie-dlia-masshtabirovaniia-na-etape-testirovaniia/2026-06-18weekly0.7 https://korshunov.ai/en/article/1563-llms-struggle-with-negation-in-figurative-language/2026-06-18weekly0.7 https://korshunov.ai/en/article/1559-graphpo-graph-based-policy-optimization-for-reasoning-models/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1561-senflow-rasshirennaia-detektsiia-teksta-sgenerirovannogo-ii-v-gibridnykh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1561-senflow-advanced-ai-generated-text-detection-in-hybrid-documents/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1560-razdelenie-poiska-ot-rassuzhdenii-v-agentakh-na-osnove-iazykovykh-modelei/2026-06-18weekly0.7 https://korshunov.ai/en/article/1560-decoupling-search-from-reasoning-in-llm-agents/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1559-graphpo-grafovaia-optimizatsiia-politiki-dlia-modelei-rassuzhdeniia/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1558-otsenka-dementsii-na-osnove-rechi-s-kompensatsiei-oshibok/2026-06-18weekly0.7 https://korshunov.ai/en/article/1556-g-idiomalign-gloss-pivoted-benchmark-for-cross-lingual-idiom-alignment/2026-06-18weekly0.7 https://korshunov.ai/en/article/1558-speech-based-dementia-assessment-with-error-mitigation/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1556-g-idiomalign-benchmark-s-ispolzovaniem-glagolnogo-opornogo-elementa-dlia-idiom/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1557-cade-priamoe-vstraivanie-vremennykh-shagov-dlia-otvetov-na-voprosy-po-vremennym/2026-06-18weekly0.7 https://korshunov.ai/en/article/1557-cade-direct-timestep-embedding-for-time-series-question-answering/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1552-srednie-i-pozdnie-chasti-nauchnykh-statei-raskryvaiut-kliuchevuiu-informatsiiu/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1555-reguliruemoe-sliianie-modelei-dlia-mnogoznachnogo-myshleniia/2026-06-18weekly0.7 https://korshunov.ai/en/article/1555-steerable-model-merging-for-multilingual-reasoning/2026-06-18weekly0.7 https://korshunov.ai/en/article/1552-middle-to-late-segments-of-research-papers-reveal-key-methodological-information/2026-06-18weekly0.7 https://korshunov.ai/en/article/1554-sumi-open-uniform-diffusion-language-model-from-scratch/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1554-sumi-otkrytaia-unifitsirovannaia-model-raspredelennoi-generatsii-iazyka-s-nulia/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1547-ustranenie-bolezni-indeksa-za-schet-fizicheskoi-razdelnosti-bazovoi-i/2026-06-18weekly0.7 https://korshunov.ai/en/article/1546-urdu-katib-handwritten-dataset-released-for-uhtr-research/2026-06-18weekly0.7 https://korshunov.ai/en/article/1550-leadership-as-coordination-control-in-multi-agent-llm-teams/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1550-vedenie-kak-koordinatsionnoe-upravlenie-v-multiagentnykh-komandakh-na-osnove-llm/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1546-vypushchen-nabor-dannykh-po-ruchnomu-pismu-na-urdu-dlia-issledovanii-uhtr/2026-06-18weekly0.7 https://korshunov.ai/en/article/1547-index-sickness-elimination-via-baseline-log-physical-separation/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1541-kadrovyi-freimvork-chelovecheskoi-iskusstvennoi-intellekt-sovmestnoi-evoliutsii/2026-06-18weekly0.7 https://korshunov.ai/en/article/1532-leaked-financial-docs-show-openai-is-losing-billions-of-dollars-a-year/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1532-potekshie-finansovye-dokumenty-pokazyvaiut-chto-openai-teriaet-milliardy-v-god/2026-06-18weekly0.7 https://korshunov.ai/en/article/1538-indiccontexteval-benchmark-for-context-utilisation-in-audio-llms/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1538-indiccontexteval-benchmark-dlia-ispolzovaniia-konteksta-v-audio-modeliakh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1541-human-ai-coevolution-framework-reveals-social-intelligence-emergence/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1527-kontrol-obratnoi-sviazi-pid-dlia-interpretiruemogo-upravleniia-aktivatsiei-v/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1526-shift-umenshenie-iazykovoi-predvziatosti-v-multiazhnykh-sistemakh-poiska/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1528-r2d-rl-sreda-2d-futbola-robocup-dlia-marl/2026-06-18weekly0.7 https://korshunov.ai/en/article/1527-pid-feedback-control-for-interpretable-activation-steering-in-music-generation/2026-06-18weekly0.7 https://korshunov.ai/en/article/1528-r2d-rl-robocup-2d-soccer-environment-for-marl/2026-06-18weekly0.7 https://korshunov.ai/en/article/1526-shift-reducing-language-bias-in-multilingual-information-retrieval/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1524-samouslovlennoe-prisvoenie-kreditov-dlia-rl-s-podtverzhdaemymi/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1525-profillm-profilirovanie-polzovatelei-s-orientatsiei-na-poleznost-dlia-zakazov-v/2026-06-18weekly0.7 https://korshunov.ai/en/article/1525-profillm-utility-aligned-agentic-user-profiling-for-industrial-ride-hailing/2026-06-18weekly0.7 https://korshunov.ai/en/article/1524-self-conditioned-credit-assignment-for-rl-with-verifiable-rewards/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1523-reskaling-mlm-golovy-dlia-neirospartogo-poiska/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1521-switchbraidnet-legkovesnaia-model-eeg-dlia-gibridnykh-bis/2026-06-18weekly0.7 https://korshunov.ai/en/article/1521-switchbraidnet-lightweight-eeg-model-for-hybrid-bcis/2026-06-18weekly0.7 https://korshunov.ai/en/article/1523-rescaling-mlm-head-for-neural-sparse-retrieval/2026-06-18weekly0.7 https://korshunov.ai/en/article/1522-reinforcement-learning-foundation-models-should-already-be-a-thing/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1522-fundamentalnye-modeli-obucheniia-s-usileniem-dolzhny-byt-uzhe/2026-06-18weekly0.7 https://korshunov.ai/en/article/1520-maturing-markov-decision-processes-introduce-new-decision-framework/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1520-sozrevaiushchie-markovskie-protsessy-priniatiia-reshenii-vvodiat-novyi-freimvork/2026-06-18weekly0.7 https://korshunov.ai/en/article/1518-data-recipe-boosts-long-context-reasoning-in-llms/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1518-dannye-retsept-uluchshaet-dolgosrochnoe-myshlenie-v-bolshikh-iazykovykh/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1519-prostranstvo-eto-intellekt-neironnaia-semigruppovaia-superpozitsiia-dlia-metriki/2026-06-18weekly0.7 https://korshunov.ai/en/article/1519-space-is-intelligence-neural-semigroup-superposition-for-riemannian-metric/2026-06-18weekly0.7 https://korshunov.ai/en/article/1517-trust-target-confidence-recourse-with-tsetlin-machines/2026-06-18weekly0.7 https://korshunov.ai/en/article/1516-robot-uses-prior-team-experience-to-improve-usar-rescue-success/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1516-robot-ispolzuet-opyt-predydushchei-komandy-dlia-uluchsheniia-uspekha-operatsii-v/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1515-skill-mas-evoliutsionnaia-metaskill-dlia-avtomaticheskikh-sistem-mnogoagentov/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1517-trust-metod-s-uchetom-doveriia-s-ispolzovaniem-mashin-tsetlin/2026-06-18weekly0.7 https://korshunov.ai/en/article/1515-skill-mas-evolving-meta-skill-for-automatic-multi-agent-systems/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1512-kinemaforge-sintez-urdf-iz-posledovatelnostei-rgb-d/2026-06-18weekly0.7 https://korshunov.ai/en/article/1512-kinemaforge-urdf-synthesis-from-rgb-d-sequences/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1513-impsh-uluchshaet-obnaruzhenie-skrytogo-repressivnogo-soderzhaniia-v-raznykh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1514-worldlines-benchmarking-long-horizon-embodied-agent-memory/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1514-worldlines-otsenka-pamiati-agentov-s-dolgosrochnymi-gorizontami-v-usloviiakh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1513-impsh-improves-implicit-hate-speech-detection-across-domains/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1509-neironnye-seti-s-uchetom-sdviga-domena-dlia-otsenki-nesbalansirovannykh-mass/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1511-masshtabirovanie-sistemy-avariinogo-tormozheniia-s-ispolzovaniem-ogromnykh-meta/2026-06-18weekly0.7 https://korshunov.ai/en/article/1509-domain-shift-aware-neural-networks-for-unbalance-mass-estimation/2026-06-18weekly0.7 https://korshunov.ai/en/article/1510-xcientist-externalizing-research-synthesis-and-validation-in-ai-scientists/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1510-xcientist-vneshniaia-sintezatsiia-i-proverka-issledovanii-v-ii-issledovateliakh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1511-scaling-aeb-with-massive-unlabeled-data-via-meta-feedback-ssl/2026-06-18weekly0.7 https://korshunov.ai/en/article/1507-skill-guided-continuation-distillation-for-gui-agents/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1507-navyk-orientirovannaia-distilliatsiia-dlia-agentov-interfeisa/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1508-beliefdiffusion-generativnoe-predskazatelnoe-planirovanie-dlia-navigatsii/2026-06-18weekly0.7 https://korshunov.ai/en/article/1508-beliefdiffusion-generative-model-predictive-planning-for-navigation/2026-06-18weekly0.7 https://korshunov.ai/en/article/1506-saerec-fine-grained-intent-priors-via-sparse-autoencoders/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1506-saerec-chastotnye-predposylki-namerenii-cherez-razrezhennye-avtoenkodery/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1503-scirisk-bench-benchmark-orientirovannyi-na-otsenku-riskov-v-oblasti-bezopasnosti/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1504-transitnet-dostigaet-tochnosti-952-pri-poiske-tranzitov-v-usloviiakh-nizkogo/2026-06-18weekly0.7 https://korshunov.ai/en/article/1505-llms-struggle-with-negation-in-figurative-language/2026-06-18weekly0.7 https://korshunov.ai/en/article/1503-scirisk-bench-a-risk-dimension-aware-benchmark-for-ai4science-safety/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1505-modeli-llm-ispytyvaiut-trudnosti-s-otritsaniem-v-metaforicheskoi-rechi/2026-06-18weekly0.7 https://korshunov.ai/en/article/1504-transitnet-achieves-95-2-accuracy-in-low-snr-transit-searches/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1500-kvantovaia-augmentatsiia-gan-ne-pokazyvaet-preimushchestv-v-mrt-mozga/2026-06-18weekly0.7 https://korshunov.ai/en/article/1500-quantum-gan-augmentation-shows-no-benefit-in-brain-mri/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1502-razdelenie-poiska-ot-rassuzhdenii-v-agentakh-na-osnove-iazykovykh-modelei/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1501-rtsgamebench-benchmark-dlia-strategicheskogo-myshleniia-v-igrakh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1501-rtsgamebench-an-rts-benchmark-for-strategic-reasoning/2026-06-18weekly0.7 https://korshunov.ai/en/article/1502-decoupling-search-from-reasoning-in-llm-agents/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1497-thinkdeception-interpretiruemaia-mnogomodalnaia-sistema-obnaruzheniia-lzhi/2026-06-18weekly0.7 https://korshunov.ai/en/article/1499-capra-multi-agent-llm-system-for-software-architecture-feedback/2026-06-18weekly0.7 https://korshunov.ai/en/article/1497-thinkdeception-interpretable-multimodal-deception-detection-framework/2026-06-18weekly0.7 https://korshunov.ai/en/article/1498-cade-direct-timestep-embedding-for-time-series-question-answering/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1498-cade-priamoe-vstraivanie-vremennykh-shagov-dlia-otvetov-na-voprosy-po-vremennym/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1499-capra-mnogoagentnaia-sistema-llm-dlia-obratnoi-sviazi-po-arkhitekture/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1496-g-idiomalign-benchmark-s-ispolzovaniem-glagolov-pivotov-dlia-sinkhronizatsii/2026-06-18weekly0.7 https://korshunov.ai/en/article/1496-g-idiomalign-gloss-pivoted-benchmark-for-cross-lingual-idiom-alignment/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1495-trap-benchmark-dlia-vypolneniia-zadach-i-ustoichivosti-k-aktivnomu-izvlecheniiu/2026-06-18weekly0.7 https://korshunov.ai/en/article/1495-trap-benchmark-for-task-completion-and-resistance-to-active-privacy-extraction/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1494-fokus-ispolzovanie-gpu-spot-dlia-uskoreniia-postobuchivaniia-dit-rl/2026-06-18weekly0.7 https://korshunov.ai/en/article/1494-spotlight-using-spot-gpus-to-accelerate-dit-rl-post-training/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1491-izmenchivost-v-programmnom-obespechenii-sozdannom-s-pomoshchiu-ii-novyi-podkhod/2026-06-18weekly0.7 https://korshunov.ai/en/article/1492-lstm-vision-transformer-improves-hrrr-forecast-error-prediction/2026-06-18weekly0.7 https://korshunov.ai/en/article/1493-fomoe-breaks-full-replica-barrier-with-partitioned-expert-layers/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1492-lstm-videnie-transformer-uluchshaet-predskazanie-oshibok-prognoza-hrrr/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1493-fomoe-preodolevaet-porog-polnykh-kopii-s-pomoshchiu-razdelennykh-ekspertiz-sloev/2026-06-18weekly0.7 https://korshunov.ai/en/article/1491-variability-in-ai-generated-software-a-new-product-line-approach/2026-06-18weekly0.7 https://korshunov.ai/en/article/1489-ariadne-agnostic-routing-for-inference-time-adapter-selection/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1489-ariadne-agnostichnoe-marshrutizatsiia-dlia-vybora-adapterov-na-etape-inferensa/2026-06-18weekly0.7 https://korshunov.ai/en/article/1488-productconsistency-enhancing-product-identity-in-image-editing/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1490-rods-nagrada-orientirovannoe-onlain-sintezirovanie-dannykh-dlia-agentov-s/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1488-productconsistency-uluchshenie-identichnosti-produkta-v-redaktirovanii/2026-06-18weekly0.7 https://korshunov.ai/en/article/1490-rods-reward-driven-online-data-synthesis-for-multi-turn-tool-use-agents/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1487-vedenie-kak-koordinatsionnyi-kontrol-v-mnogoagentnykh-komandakh-na-osnove-llm/2026-06-18weekly0.7 https://korshunov.ai/en/article/1487-leadership-as-coordination-control-in-multi-agent-llm-teams/2026-06-18weekly0.7 https://korshunov.ai/en/article/1486-towards-an-agent-first-web-redesigning-the-web-for-ai-agents/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1486-k-veb-srede-orientirovannoi-na-agentov-perestroika-veba-dlia-iskusstvennykh/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1485-xai-raskryvaet-kliuchevye-faktory-v-evropeiskikh-elektricheskikh-rynkakh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1485-xai-reveals-key-drivers-in-european-electricity-markets/2026-06-18weekly0.7 https://korshunov.ai/en/article/1484-equivariant-graph-neural-networks-improve-optical-spectra-prediction/2026-06-18weekly0.7 https://korshunov.ai/en/article/1483-pareto-q-learning-with-reward-machines/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1483-pareto-q-learning-s-mashinami-voznagrazhdenii/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1482-tekhnicheskaia-klassifikatsiia-protokolov-kommunikatsii-agentov-na-osnove-llm/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1484-ekvivariantnye-grafovye-neironnye-seti-uluchshaiut-predskazanie-opticheskikh/2026-06-18weekly0.7 https://korshunov.ai/en/article/1482-technical-taxonomy-of-llm-agent-communication-protocols/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1480-orthoreg-ortogonalnaia-reguliarizatsiia-dlia-gibridnykh-simvolicheskikh-sistem/2026-06-18weekly0.7 https://korshunov.ai/en/article/1479-adsmind-physics-grounded-multi-agent-system-for-adsorption-discovery/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1481-kadrovaia-ramka-sovmestnogo-evoliutsionnogo-razvitiia-cheloveka-i-ii-raskryvaet/2026-06-18weekly0.7 https://korshunov.ai/en/article/1480-orthoreg-orthogonal-regularization-for-hybrid-symbolic-neural-dynamical-systems/2026-06-18weekly0.7 https://korshunov.ai/en/article/1481-human-ai-coevolution-framework-reveals-social-intelligence-emergence/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1479-adsmind-fizicheski-obosnovannaia-mnogopolzovatelskaia-sistema-dlia-otkrytiia/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1474-glm-5-2-vedushchaia-model-otkrytykh-vesov-na-indekse-iskusstvennogo-intellekta/2026-06-18weekly0.7 https://korshunov.ai/en/article/1474-glm-5-2-is-the-leading-open-weights-model-on-the-artificial-analysis-index/2026-06-18weekly0.7 https://korshunov.ai/ru/article/1471-llama-cpp-teper-podderzhivaet-upravlenie-modeliami-cherez-api/2026-06-17weekly0.7 https://korshunov.ai/en/article/1470-i-released-inflect-nano-an-ultra-extreme-tiny-4-63m-parameter-tts-model/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1470-ia-vypustil-inflect-nano-ekstremalno-malenkii-model-tts-s-463-mln-parametrov/2026-06-17weekly0.7 https://korshunov.ai/en/article/1471-llama-cpp-now-supports-model-management-via-api/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1466-laboratoriia-lin-tsziunia-po-ii-zavershaet-raund-finansirovaniia-na-otsenke-v-2/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1469-luchshie-modeli-dlia-videopamiati-12-gb/2026-06-17weekly0.7 https://korshunov.ai/en/article/1466-lin-junyang-ai-lab-closes-2b-valuation-round/2026-06-17weekly0.7 https://korshunov.ai/en/article/1469-best-models-for-a-12gb-vram-card/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1467-ofitsialnye-zametki-po-vypusku-claude-code-v2-1-181/2026-06-17weekly0.7 https://korshunov.ai/en/article/1467-claude-code-v2-1-181-release-notes/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1460-ggml-cpu-vkliuchat-bekend-power11-uslovno-v-zavisimosti-ot-podderzhki/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1459-llama-cpp-release-b9692-adds-new-binaries-and-fixes/2026-06-17weekly0.7 https://korshunov.ai/en/article/1459-llama-cpp-release-b9692-adds-new-binaries-and-fixes/2026-06-17weekly0.7 https://korshunov.ai/en/article/1460-ggml-cpu-conditionally-enable-power11-backend-based-on-compiler-support/2026-06-17weekly0.7 https://korshunov.ai/en/article/1453-lemonade-v10-8-releases-auto-memory-management-cloud-offload-and-mcp-tool/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1453-lemonade-v10-8-vypuskaet-avtomaticheskoe-upravlenie-pamiatiu-podderzhku-oblaka-i/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1452-video-o-vypuske-glm-5-2-sozdano-s-ispolzovaniem-glm-5-2/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1451-bekend-metal-dobavliaet-podderzhku-f16-i-bf16-dlia-operatora-concat/2026-06-17weekly0.7 https://korshunov.ai/en/article/1452-glm-5-2-release-video-made-with-glm-5-2/2026-06-17weekly0.7 https://korshunov.ai/en/article/1451-metal-backend-adds-f16-and-bf16-support-for-concat-operator/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1449-my-srochno-nuzhdaemsia-v-modeli-80-160b-dlia-ustroistv-s-edinoi-pamiatiu/2026-06-17weekly0.7 https://korshunov.ai/en/article/1450-llama-cpp-releases-b9690-with-rope-back-operator-and-cross-platform-binaries/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1450-llama-cpp-vypuskaet-b9690-s-operatorom-rope-back-i-binarnikami-dlia-raznykh/2026-06-17weekly0.7 https://korshunov.ai/en/article/1449-we-need-a-80-160b-model-urgently-for-unified-memory-devices/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1446-ssha-otlozhili-vkliuchenie-v-chernyi-spisok-kitaiskoi-kompanii-deepseek/2026-06-17weekly0.7 https://korshunov.ai/en/article/1447-no-jailbreak-fable-s-fix-this-code-was-a-fake-scenario/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1447-bez-proniknoveniia-stsenarii-ispravte-etot-kod-v-fable-byl-falshivym/2026-06-17weekly0.7 https://korshunov.ai/en/article/1446-us-holds-off-blacklisting-china-s-deepseek/2026-06-17weekly0.7 https://korshunov.ai/en/article/1443-llama-cpp-releases-version-b9688-with-new-apis-and-cross-platform-binaries/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1444-llama-cpp-release-b9687-adds-new-binaries-and-fixes/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1443-llama-cpp-vypuskaet-versiiu-b9688-s-novymi-api-i-binarnikami-dlia-raznykh/2026-06-17weekly0.7 https://korshunov.ai/en/article/1444-llama-cpp-release-b9687-adds-new-binaries-and-fixes/2026-06-17weekly0.7 https://korshunov.ai/en/article/1440-radical-ai-achieves-10x-acceleration-in-materials-discovery/2026-06-17weekly0.7 https://korshunov.ai/en/article/1442-i-post-trained-a-model-to-reliably-roll-a-die/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1442-ia-poobuchil-model-dlia-nadezhnogo-broska-kubika/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1441-psa-unsloth-glm-5-2-gguf-nakhoditsia-v-protsesse-zagruzki/2026-06-17weekly0.7 https://korshunov.ai/en/article/1441-psa-unsloth-glm-5-2-gguf-is-uploading/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1440-radical-ai-dostigaet-uskoreniia-v-10-raz-v-otkrytii-materialov/2026-06-17weekly0.7 https://korshunov.ai/en/article/1439-llama-cpp-gpu-memory-optimization-tips/2026-06-17weekly0.7 https://korshunov.ai/en/article/1437-we-built-an-open-source-ui-kit-for-document-rag-agents/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1437-my-sozdali-otkrytyi-istochnik-ui-kit-dlia-rag-agentov-dokumentov/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1439-sovety-po-optimizatsii-pamiati-gpu-dlia-llama-cpp/2026-06-17weekly0.7 https://korshunov.ai/en/article/1438-glm-5-2-fp8-hgx-h200-sglang-docker-deployment-config/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1438-glm-5-2-fp8-hgx-h200-sglang-docker-deployment-config/2026-06-17weekly0.7 https://korshunov.ai/en/article/1431-llama-cpp-release-b9686-fixes-segfault-on-long-prompts-for-eagle3/2026-06-17weekly0.7 https://korshunov.ai/en/article/1432-llama-cpp-release-b9685-adds-sycl-dev2dev-memcpy-and-multiple-platform-binaries/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1432-reliz-llama-cpp-b9685-dobavliaet-sycl-dev2dev-memcpy-i-neskolko-platformennykh/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1431-llama-cpp-release-b9686-ispravlenie-segfolta-pri-dlinnykh-zaprosakh-dlia-eagle3/2026-06-17weekly0.7 https://korshunov.ai/en/article/1428-gemma-4-e2b-runs-at-255-tok-s-in-browser-using-webgpu/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1428-gemma-4-e2b-rabotaet-so-skorostiu-255-tokenov-v-sekundu-v-brauzere-s-webgpu/2026-06-17weekly0.7 https://korshunov.ai/en/article/1429-gamecraft-bench-can-agents-build-playable-games-end-to-end-in-a-real-game-engine/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1429-gamecraft-bench-mozhet-li-agent-sozdavat-razvertyvaemye-igry-v-realnom-dvizhke/2026-06-17weekly0.7 https://korshunov.ai/en/article/1430-loopcoder-v2-two-loop-plt-model-achieves-best-gain-cost-trade-off/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1430-loopcoder-v2-model-plt-s-dvumia-tsiklami-dostigaet-nailuchshego-sootnosheniia-i/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1422-ii-trebuet-bolshei-inzhenernoi-distsipliny/2026-06-17weekly0.7 https://korshunov.ai/en/article/1422-ai-demands-more-engineering-discipline/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1421-kto-to-eshche-ne-mozhet-naiti-veb-interfeis-v-poslednei-versii-llama-cpp-b9680/2026-06-17weekly0.7 https://korshunov.ai/en/article/1421-is-anyone-else-not-finding-the-web-ui-on-latest-llama-cpp-b9680/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1419-reliz-llama-cpp-b9684-dobavliaet-conv-3d-i-binarniki-dlia-neskolkikh-platform/2026-06-17weekly0.7 https://korshunov.ai/en/article/1419-llama-cpp-release-b9684-adds-conv-3d-and-multiple-platform-binaries/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1413-sozdanie-biudzhetnykh-modelei-kotorye-prevoskhodiat-svoi-razmery-blagodaria-na/2026-06-17weekly0.7 https://korshunov.ai/en/article/1413-making-budget-models-punch-above-their-weight-with-a-smart-rust-harness/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1412-trellis-2-teper-rabotaet-nativno-na-mlx/2026-06-17weekly0.7 https://korshunov.ai/en/article/1412-trellis-2-now-runs-natively-on-mlx/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1410-glm-5-2-na-4x-sparks-razumno/2026-06-17weekly0.7 https://korshunov.ai/en/article/1410-glm-5-2-on-4x-sparks-reasonable/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1403-glm-5-2-pobeda-dlia-lokalnoi-ii/2026-06-17weekly0.7 https://korshunov.ai/en/article/1404-llama-cpp-release-b9682-adds-vulkan-support-and-new-platform-binaries/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1404-llama-cpp-release-b9682-dobavliaet-podderzhku-vulkan-i-novye-binarniki-dlia/2026-06-17weekly0.7 https://korshunov.ai/en/article/1405-state-of-the-interconnects-blog-mid-2026/2026-06-17weekly0.7 https://korshunov.ai/en/article/1403-glm-5-2-is-a-win-for-local-ai/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1405-blog-o-sostoianii-soedinitelnykh-tekhnologii-seredina-2026-goda/2026-06-17weekly0.7 https://korshunov.ai/en/article/1400-headless-screenshot-loops-enable-a-30b-local-agent-to-debug-raytraced-fps-in-c/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1400-tsikly-bezdeistvuiushchikh-skrinshotov-pozvoliaiut-30b-lokalnomu-agentu-otladit/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1395-reliz-llama-cpp-b9678-dobavliaet-optimizatsii-i-sborki-dlia-neskolkikh-platform/2026-06-17weekly0.7 https://korshunov.ai/en/article/1398-local-llm-powered-rpg-with-persistent-generated-content/2026-06-17weekly0.7 https://korshunov.ai/en/article/1399-siq-1-qwen3-6-achieves-strong-performance-in-autoresearch-and-benchmarking/2026-06-17weekly0.7 https://korshunov.ai/en/article/1395-llama-cpp-release-b9678-adds-optimizations-and-cross-platform-builds/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1398-lokalnaia-rpg-igra-s-ispolzovaniem-lokalnykh-modelei-iazykovoi-obrabotki-s/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1399-siq-1-qwen3-6-dostigaet-silnykh-rezultatov-v-avtoznaniiakh-i-testirovanii/2026-06-17weekly0.7 https://korshunov.ai/en/article/1396-llama-cpp-release-b9677-updates-and-cross-platform-binaries/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1396-llama-cpp-release-b9677-obnovleniia-i-binarniki-dlia-raznykh-platform/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1393-reshena-li-zadacha-poiska-igolki-v-stoge-sena/2026-06-17weekly0.7 https://korshunov.ai/en/article/1393-is-the-needle-in-haystack-problem-solved/2026-06-17weekly0.7 https://korshunov.ai/en/article/1390-llama-cpp-release-b9675-adds-fp16-support-and-new-platform-binaries/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1391-reliz-llama-cpp-b9674-ispravlenie-baga-s-asinkhronnym-memcpy-i-dobavlenie-novykh/2026-06-17weekly0.7 https://korshunov.ai/en/article/1391-llama-cpp-release-b9674-fixes-async-memcpy-bug-and-adds-new-binaries/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1390-llama-cpp-release-b9675-dobavliaet-podderzhku-fp16-i-novye-binarniki-dlia/2026-06-17weekly0.7 https://korshunov.ai/en/article/1387-local-models-went-from-mostly-useless-to-actually-useful-in-one-year/2026-06-17weekly0.7 https://korshunov.ai/en/article/1388-best-estimate-on-how-long-we-will-see-fable-5-class-open-weight-models/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1388-luchshaia-otsenka-po-srokam-v-techenie-kotorykh-my-uvidim-modeli-otkrytogo-vesa/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1387-lokalnye-modeli-pereshli-ot-v-osnovnom-bespoleznykh-do-deistvitelno-poleznykh-za/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1381-glm-5-2-sozdan-dlia-zadach-s-dlitelnym-gorizontom/2026-06-17weekly0.7 https://korshunov.ai/en/article/1382-a-year-building-a-fully-local-home-voice-assistant/2026-06-17weekly0.7 https://korshunov.ai/en/article/1381-glm-5-2-built-for-long-horizon-tasks/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1382-god-razrabotki-polnostiu-lokalnogo-golosovogo-pomoshchnika-dlia-doma/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1379-llama-cpp-release-b9680-new-binaries-and-vulkan-support/2026-06-17weekly0.7 https://korshunov.ai/en/article/1379-llama-cpp-release-b9680-new-binaries-and-vulkan-support/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1378-rio-3-5-397b-veroiatno-neudachnoe-prestuplenie-po-krazhe-sredstv-na-razvitie-ii/2026-06-17weekly0.7 https://korshunov.ai/en/article/1378-rio-3-5-397b-likely-a-failed-embezzlement-of-ai-funding/2026-06-17weekly0.7 https://korshunov.ai/en/article/1374-llama-cpp-releases-b9673-with-usm-system-allocations-and-cross-platform-binaries/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1374-llama-cpp-vypuskaet-b9673-s-allokatsiiami-usm-sistemy-i-binarnikami-dlia-raznykh/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1373-glm-5-2-max-iavliaetsia-tekushchim-tretim-po-velichine-modeliu/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1369-datasette-1-0a34-dobavliaet-instrumenty-redaktirovaniia-i-udaleniia-strok/2026-06-17weekly0.7 https://korshunov.ai/en/article/1369-datasette-1-0a34-adds-row-editing-and-deletion-tools/2026-06-17weekly0.7 https://korshunov.ai/en/article/1373-glm-5-2-max-is-currently-the-third-best-model/2026-06-17weekly0.7 https://korshunov.ai/en/article/1371-looking-for-locally-hosted-tool-to-create-english-subtitles-from-videos/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1371-poisk-lokalno-razmeshchaemogo-instrumenta-dlia-sozdaniia-angliiskikh-subtitrov/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1367-click-to-play-izobrazhenie-kotoroe-vosproizvoditsia-pri-klike/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1366-glm-5-2-pretenduet-na-pervoe-mesto-v-frontend-programmirovanii-s-ispolzovaniem/2026-06-17weekly0.7 https://korshunov.ai/en/article/1366-glm-5-2-claims-top-position-in-frontend-coding-with-speculative-decoding/2026-06-17weekly0.7 https://korshunov.ai/en/article/1367-click-to-play-a-still-that-plays/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1364-sravnenie-proizvoditelnosti-rtx-5060-ti-16gb-i-rx-9060-xt-16gb/2026-06-17weekly0.7 https://korshunov.ai/en/article/1364-rtx-5060-ti-16gb-vs-rx-9060-xt-16gb-benchmark-comparison/2026-06-17weekly0.7 https://korshunov.ai/en/article/1362-elias-in-the-lighthouse-diagnosing-low-diversity-in-llm-stories/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1362-elias-v-fonare-diagnostika-nizkoi-raznoobraznosti-v-istoriiakh-bolshikh-modelei/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1360-legalhallulens-audit-khalliutsinatsii-v-pravovykh-ii/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1359-rekursivnye-skrytye-modeli-diffuzii-vvodiat-novyi-osevoi-masshtab/2026-06-17weekly0.7 https://korshunov.ai/en/article/1359-recursive-masked-diffusion-models-introduce-new-scaling-axis/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1358-loopcoder-v2-dostigaet-optimalnoi-proizvoditelnosti-pri-dvukh-tsiklakh/2026-06-17weekly0.7 https://korshunov.ai/en/article/1358-loopcoder-v2-achieves-optimal-two-loop-performance/2026-06-17weekly0.7 https://korshunov.ai/en/article/1360-legalhallulens-auditing-hallucinations-in-legal-ai/2026-06-17weekly0.7 https://korshunov.ai/en/article/1339-baseline-evaluation-of-open-source-llms-for-multi-label-att-ck-classification/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1357-katastroficheskoe-zabvenie-eto-nizkii-rang-teoriia-funktsionalnogo-prostranstva/2026-06-17weekly0.7 https://korshunov.ai/en/article/1357-catastrophic-forgetting-is-low-rank-a-function-space-theory/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1356-ini-vpinn-neironnaia-set-s-fizicheskimi-usloviiami-s-uproshchennym-obrabotkoi/2026-06-17weekly0.7 https://korshunov.ai/en/article/1356-ini-vpinn-physics-informed-neural-network-with-implicit-boundary-handling/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1339-otsenka-bazovykh-otkrytykh-modelei-iazykovoi-modeli-dlia-mnogoznachnoi-att-ck/2026-06-17weekly0.7 https://korshunov.ai/en/article/1354-contex-global-counterfactual-generation-for-time-series-forecasting/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1355-otsenka-neopredelennosti-dlia-modelei-na-osnove-potokov-vizualno-iazykovykh/2026-06-17weekly0.7 https://korshunov.ai/en/article/1355-uncertainty-quantification-for-flow-based-vision-language-action-models/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1354-contex-globalnoe-generirovanie-kontraprotivnykh-stsenariev-dlia-prognozirovaniia/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1353-scafe-ispolzovanie-llm-dlia-izvlecheniia-klinicheski-znachimykh-priznakov/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1352-noisetilt-noise-tilted-reverse-kernels-dlia-vyravnivaniia-voznagrazhdeniia-v/2026-06-17weekly0.7 https://korshunov.ai/en/article/1352-noisetilt-noise-tilted-reverse-kernels-for-diffusion-reward-alignment/2026-06-17weekly0.7 https://korshunov.ai/en/article/1353-scafe-using-llms-to-extract-clinically-meaningful-scar-features/2026-06-17weekly0.7 https://korshunov.ai/en/article/1350-tensor-based-second-order-causal-discovery-algorithm/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1351-modeli-volterry-vvodiat-drobnuiu-shum-dlia-generatsii-na-osnove-scheta/2026-06-17weekly0.7 https://korshunov.ai/en/article/1351-volterra-generative-models-introduce-fractional-noise-for-score-based-generation/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1350-algoritm-vtorogo-poriadka-poiska-prichinnosti-na-osnove-tenzorov/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1348-sostavnaia-obobshchenie-v-ratsionalnykh-protsessakh-iazykovykh-modelei/2026-06-17weekly0.7 https://korshunov.ai/en/article/1349-edge-flow-a-continuous-time-model-for-gradient-descent-at-edge-of-stability/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1349-edge-flow-trekhmernaia-model-dlia-gradientnogo-spuska-na-granitse-stabilnosti/2026-06-17weekly0.7 https://korshunov.ai/en/article/1348-compositional-generalization-in-language-model-reasoning/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1347-s4op-metod-priamogo-sokrashcheniia-na-urovne-operatorov-dlia-effektivnoi-ssm/2026-06-17weekly0.7 https://korshunov.ai/en/article/1347-s4op-operator-level-pruning-for-efficient-ssm-deployment/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1345-glubokoe-obuchenie-s-podkrepleniem-dlia-minimalnogo-nulia-vynuzhdaiushchego/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1346-omniplan-adaptivnaia-sistema-dlia-svoevremennogo-i-blizkogo-optimalnogo-seti/2026-06-17weekly0.7 https://korshunov.ai/en/article/1346-omniplan-adaptive-framework-for-timely-and-near-optimal-network-planning/2026-06-17weekly0.7 https://korshunov.ai/en/article/1345-deep-reinforcement-learning-for-minimum-zero-forcing-sets/2026-06-17weekly0.7 https://korshunov.ai/en/article/1344-learning-fair-pareto-optimal-policies-in-multi-objective-reinforcement-learning/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1343-ternarnyi-mamba-effektivnaia-qat-modelei-ssm-iz-predvaritelno-obuchennykh/2026-06-17weekly0.7 https://korshunov.ai/en/article/1343-ternary-mamba-efficient-qat-of-ssms-from-pretrained-checkpoints/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1344-izuchenie-spravedlivykh-pareto-optimalnykh-politik-v-mnogokriterialnoi-robastnoi/2026-06-17weekly0.7 https://korshunov.ai/en/article/1341-embedded-ml-workflow-for-microcontroller-edge-devices/2026-06-17weekly0.7 https://korshunov.ai/en/article/1342-handlebars-triple-brace-injection-exploits-structural-role-delimiters/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1342-narushenie-vvoda-s-pomoshchiu-troinykh-figurnykh-skobok-v-handlebars-ispolzovat/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1341-vstroennaia-rabochaia-sreda-ml-dlia-periferiinykh-ustroistv-na/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1338-lil-q-konveksnyi-metod-dlia-nelineinykh-differentsialnykh-uravnenii-v-chastnykh/2026-06-17weekly0.7 https://korshunov.ai/en/article/1338-lil-q-convex-method-for-nonlinear-pdes-with-pinns/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1340-prodolzhitelnost-raboty-kak-umenshaiushchiisia-kapital-v-pamiati-robota/2026-06-17weekly0.7 https://korshunov.ai/en/article/1340-flash-endurance-as-depreciating-capital-in-robot-memory/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1337-diffuzionnoe-priblizhenie-dlia-obucheniia-td-s-lineinymi-priznakami/2026-06-17weekly0.7 https://korshunov.ai/en/article/1336-kolmogorov-regression-for-robust-diffusion-policies/2026-06-17weekly0.7 https://korshunov.ai/en/article/1335-att-ck-labeled-multi-source-cybersecurity-logs-dataset-released/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1335-vypushchena-baza-dannykh-s-metkami-att-ck-obediniaiushchaia-logi-iz-razlichnykh/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1336-kolmogorovskaia-regressiia-dlia-ustoichivykh-diffuzionnykh-politik/2026-06-17weekly0.7 https://korshunov.ai/en/article/1337-diffusion-approximation-for-td-learning-with-linear-features/2026-06-17weekly0.7 https://korshunov.ai/en/article/1333-do-distilled-sets-outperform-coresets/2026-06-17weekly0.7 https://korshunov.ai/en/article/1332-logarithmic-scaling-of-queue-peaks-beyond-geometric-thresholds/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1332-logarifmicheskaia-shkala-pikov-ocheredei-za-geometricheskim-porogom/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1333-pobezhdaet-li-metod-szhatiia-datasetov-pered-koorsetami/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1334-modeli-tsiklicheskogo-mira-dostigaiut-effektivnosti-v-100-raz-po-parametram/2026-06-17weekly0.7 https://korshunov.ai/en/article/1334-looped-world-models-achieve-100x-parameter-efficiency/2026-06-17weekly0.7 https://korshunov.ai/en/article/1330-adavomp-adaptive-volumetric-mechanical-property-fields/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1331-izuchenie-politiki-krasnogo-agenta-na-osnove-nabliudenii-dlia-neirosimvolnykh/2026-06-17weekly0.7 https://korshunov.ai/en/article/1331-learning-red-agent-policy-from-observations-for-neurosymbolic-cyber-agents/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1330-adavomp-adaptivnye-obemnye-polia-mekhanicheskikh-svoistv/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1329-znakovyi-rang-indeks-i-povtoriaemost-spiska-sviazi-i-razryvy/2026-06-17weekly0.7 https://korshunov.ai/en/article/1329-sign-rank-index-and-list-replicability-connections-and-separations/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1328-reprorepo-masshtabiruemyi-audit-vosproizvodimosti-s-ispolzovaniem-issue-github/2026-06-17weekly0.7 https://korshunov.ai/en/article/1327-netnewswire-status-open-source-project-thrives-post-retirement/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1326-otsenki-modifitsirovannogo-v620-s-proshivkoi-w6800-na-ebay/2026-06-17weekly0.7 https://korshunov.ai/en/article/1326-benchmarks-of-modded-v620-with-w6800-firmware-on-ebay/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1327-status-netnewswire-otkrytyi-proekt-prodolzhaet-razvivatsia-posle-vykhoda-na/2026-06-17weekly0.7 https://korshunov.ai/en/article/1328-reprorepo-scalable-reproducibility-audits-with-github-issues/2026-06-17weekly0.7 https://korshunov.ai/en/article/1325-cheapest-way-to-run-glm-5-x-locally-without-unified-memory/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1325-samyi-deshevyi-sposob-zapuska-glm-5-x-lokalno-bez-ispolzovaniia-edinykh-pamiati/2026-06-17weekly0.7 https://korshunov.ai/en/article/1324-llms-predict-dementia-and-depression-from-clinical-speech/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1324-llms-predskazyvaiut-dementsiiu-i-depressiiu-na-osnove-klinicheskoi-rechi/2026-06-17weekly0.7 https://korshunov.ai/en/article/1322-source-language-effects-in-cross-lingual-in-context-learning/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1322-vliianie-iskhodnogo-iazyka-v-mezhiazykovom-obuchenii-v-kontekste/2026-06-17weekly0.7 https://korshunov.ai/en/article/1323-legalhallulens-auditing-hallucinations-in-legal-ai/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1323-legalhallulens-audit-galliutsinatsii-v-pravovom-ii/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1321-provenanceguard-proverka-dostovernosti-s-uchetom-istochnika-dlia-agentov-na-llm/2026-06-17weekly0.7 https://korshunov.ai/en/article/1321-provenanceguard-source-aware-factuality-verification-for-mcp-based-llm-agents/2026-06-17weekly0.7 https://korshunov.ai/en/article/1318-ai-s-synthetic-lived-experience-in-caregiver-support/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1318-sinteticheskii-lichnyi-opyt-ii-v-podderzhke-ukhodchika/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1319-consa-upravlenie-sparsnostiu-v-gibridnom-vnimanii/2026-06-17weekly0.7 https://korshunov.ai/en/article/1319-consa-learnable-sparsity-control-in-hybrid-attention/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1320-skillweaver-sostavnaia-marshrutizatsiia-navykov-dlia-agentov-na-osnove-modelei/2026-06-17weekly0.7 https://korshunov.ai/en/article/1320-skillweaver-compositional-skill-routing-for-llm-agents/2026-06-17weekly0.7 https://korshunov.ai/en/article/1316-security-and-privacy-prompts-in-user-llm-conversations/2026-06-17weekly0.7 https://korshunov.ai/en/article/1315-historag-integrating-historical-methodology-into-rag/2026-06-17weekly0.7 https://korshunov.ai/en/article/1317-pseudobench-benchmarking-agentic-auto-research-resistance-to-pseudoscience/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1315-historag-integratsiia-istoricheskoi-metodologii-v-rag/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1316-prompty-po-bezopasnosti-i-privatnosti-v-dialogakh-polzovatelei-s-llm/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1317-pseudobench-otsenka-ustoichivosti-agentnykh-sistem-avtomaticheskogo-poiska-k/2026-06-17weekly0.7 https://korshunov.ai/en/article/1311-measurement-gap-in-eu-law-automation/2026-06-17weekly0.7 https://korshunov.ai/en/article/1313-geographic-bias-in-large-language-models-from-user-metadata/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1313-geograficheskaia-predvziatost-v-bolshikh-iazykovykh-modeliakh-iz-metadannykh/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1314-uiazvimost-vvoda-troinykh-figurnykh-skobok-v-handlebars-pozvoliaiushchaia-roli/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1311-razryv-v-izmerenii-v-zakonodatelstve-evropeiskogo-soiuza-po-avtomatizatsii/2026-06-17weekly0.7 https://korshunov.ai/en/article/1314-handlebars-triple-brace-injection-exploits-structural-role-delimiters/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1312-otsenka-agentov-pokazyvaet-chto-modeli-ii-ne-mogut-izbegat-ekspluatatsii/2026-06-17weekly0.7 https://korshunov.ai/en/article/1312-agentic-benchmark-reveals-ai-models-fail-to-avoid-animal-exploitation/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1310-issledovanie-so-storony-krasnoi-komandy-pokazyvaet-chto-peredovye-llm-ostaiutsia/2026-06-17weekly0.7 https://korshunov.ai/en/article/1310-red-team-study-finds-frontier-llms-remain-vulnerable-to-automated-attacks/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1307-kodirovanie-slovaria-al-mawrid-s-ispolzovaniem-iso-lmf-i-tei-lex-0/2026-06-17weekly0.7 https://korshunov.ai/en/article/1307-encoding-al-mawrid-dictionary-with-iso-lmf-and-tei-lex-0/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1308-rubricstree-masshtabiruemaia-sistema-otsenki-dlia-personalnykh-agentov-zdorovia/2026-06-17weekly0.7 https://korshunov.ai/en/article/1308-rubricstree-scalable-evaluation-framework-for-personal-health-agents/2026-06-17weekly0.7 https://korshunov.ai/en/article/1309-d-opsd-on-policy-self-distillation-for-diffusion-llms/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1309-d-opsd-samodistilliatsiia-na-osnove-politiki-dlia-diffuzionnykh-llm/2026-06-17weekly0.7 https://korshunov.ai/en/article/1306-looped-world-models-achieve-100x-parameter-efficiency/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1305-zppo-uchitel-v-promptakh-a-ne-v-gradientakh/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1304-darshana-graph-korpus-dlia-sravnitelnoi-indiiskoi-filosofii/2026-06-17weekly0.7 https://korshunov.ai/en/article/1305-zppo-teacher-in-prompts-not-gradients/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1306-modeli-tsiklicheskogo-mira-dostigaiut-effektivnosti-v-100-raz/2026-06-17weekly0.7 https://korshunov.ai/en/article/1304-darshana-graph-a-corpus-for-comparative-indian-philosophy/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1300-mgup-momentum-gradientnoe-soglasovanie-dlia-selektivnoi-optimizatsii/2026-06-17weekly0.7 https://korshunov.ai/en/article/1302-variable-width-transformers-outperform-uniform-architectures/2026-06-17weekly0.7 https://korshunov.ai/en/article/1300-mgup-momentum-gradient-alignment-for-selective-optimization/2026-06-17weekly0.7 https://korshunov.ai/en/article/1303-reprorepo-scaling-reproducibility-audits-with-github-issues/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1302-peremennye-shirinnye-transformery-prevoskhodiat-odnorodnye-arkhitektury/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1303-reprorepo-masshtabirovanie-auditov-vosproizvodimosti-s-ispolzovaniem-issue/2026-06-17weekly0.7 https://korshunov.ai/en/article/1301-relar-reinforcement-guided-latent-refinement-for-stable-llm-reasoning/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1299-domen-validnost-ogranichennyi-metamorfnyi-test-dlia-surrogatov-sciml/2026-06-17weekly0.7 https://korshunov.ai/en/article/1299-domain-validity-gated-metamorphic-testing-for-sciml-surrogates/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1301-relar-regulirovanie-skrytykh-sostoianii-na-osnove-replikatsii-dlia-stabilnogo/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1298-sistema-dorozhnoi-platy-v-niu-iorke-uvelichivaet-ispolzovanie-transporta-v/2026-06-17weekly0.7 https://korshunov.ai/en/article/1298-nyc-congestion-pricing-boosts-transit-use-amid-spatially-uneven-demand-shifts/2026-06-17weekly0.7 https://korshunov.ai/en/article/1296-preference-based-trajectory-evaluation-for-agentic-systems/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1297-nmf-s-topologicheskoi-reguliarizatsiei-dlia-interpretiruemykh-bazisov/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1296-otsenka-traektorii-na-osnove-predpochtenii-dlia-agentnykh-sistem/2026-06-17weekly0.7 https://korshunov.ai/en/article/1297-nmf-with-topological-regularisation-for-interpretable-bases/2026-06-17weekly0.7 https://korshunov.ai/en/article/1295-carlos-deep-rl-for-continuous-time-optimal-stopping/2026-06-17weekly0.7 https://korshunov.ai/en/article/1293-st-cnd-framework-for-early-warning-of-geographic-tipping-points/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1294-obratnoe-obuchenie-q-novyi-algoritm-off-politseiskogo-obucheniia/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1295-carlos-glubokoe-obuchenie-dlia-nepreryvnogo-optimalnogo-ostanova/2026-06-17weekly0.7 https://korshunov.ai/en/article/1294-reversal-q-learning-a-new-off-policy-rl-algorithm/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1293-kadrovaia-sistema-st-cnd-dlia-rannego-preduprezhdeniia-geograficheskikh-tochek/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1291-scboost-umenshenie-izbytochnosti-obuchaemykh-modelei-za-schet-ortogonalizatsii/2026-06-17weekly0.7 https://korshunov.ai/en/article/1292-aoizora-topology-aware-auto-parallel-optimization-for-video-diffusion-inference/2026-06-17weekly0.7 https://korshunov.ai/en/article/1290-credit-in-event-re-anchoring-event-credit-in-dynamics-models/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1290-credit-in-event-vosstanovlenie-kreditov-sobytii-v-dinamicheskikh-modeliakh/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1292-aoizora-optimizatsiia-avtomaticheskogo-parallelizma-s-uchetom-topologii-dlia/2026-06-17weekly0.7 https://korshunov.ai/en/article/1291-scboost-reducing-learner-redundancy-via-residual-orthogonalization/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1288-selfix-kornevoi-vybor-fiksirovannogo-tochki-inversii-dlia-priamougolnykh-potokov/2026-06-17weekly0.7 https://korshunov.ai/en/article/1289-llm-features-can-hurt-gnns-via-concatenation-interference/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1289-funktsii-llm-mogut-negativno-vliiat-na-gnn-cherez-interferentsiiu-pri/2026-06-17weekly0.7 https://korshunov.ai/en/article/1287-sphere-jepa-family-of-statistical-regularizers-for-hypersphere/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1287-sphere-jepa-semeistvo-statisticheskikh-reguliatorov-dlia-gipersfery/2026-06-17weekly0.7 https://korshunov.ai/en/article/1288-selfix-root-selecting-fixed-point-inversion-for-rectified-flows-via-trajectory/2026-06-17weekly0.7 https://korshunov.ai/en/article/1286-skillmigrator-transferable-interaction-patterns-for-web-agent-efficiency/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1286-skillmigrator-perenosimye-patterny-vzaimodeistviia-dlia-effektivnosti-veb-agenta/2026-06-17weekly0.7 https://korshunov.ai/en/article/1284-physics-constrained-neural-networks-improve-weather-forecasting/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1284-fizicheski-ogranichennye-neironnye-seti-uluchshaiut-prognozirovanie-pogody/2026-06-17weekly0.7 https://korshunov.ai/en/article/1285-risk-decomposition-framework-for-pre-hoc-fine-tuning-prediction/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1285-kadrovaia-dekompozitsiia-ramki-dlia-predvaritelnoi-nastroiki-prognozirovaniia/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1282-izuchaemye-graficheskie-kuski-dlia-geterogennosti-priznakov/2026-06-17weekly0.7 https://korshunov.ai/en/article/1280-envrl-leveraging-environment-dynamics-in-agentic-rl/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1283-tuneahead-predskazyvaet-proizvoditelnost-tonkoi-nastroiki-do-nachala-obucheniia/2026-06-17weekly0.7 https://korshunov.ai/en/article/1283-tuneahead-predicts-fine-tuning-performance-before-training/2026-06-17weekly0.7 https://korshunov.ai/en/article/1282-learnable-graph-patches-for-feature-heterogeneity/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1280-envrl-ispolzovanie-dinamiki-sredy-v-agentnom-rl/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1279-spravedlivost-v-grafovykh-neironnykh-setiakh-cherez-adaptatsiiu-laplasiana/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1281-asteroid-transformator-dlia-prognozirovaniia-mnogoshagovykh-simuliatsii-dinamiki/2026-06-17weekly0.7 https://korshunov.ai/en/article/1279-fairness-in-graph-neural-networks-via-laplacian-adaptation/2026-06-17weekly0.7 https://korshunov.ai/en/article/1281-asteroid-transformer-for-multi-step-md-forecasting/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1276-modeli-vizualno-iazykovykh-sistem-ne-vsegda-nuzhdaiutsia-v-izobrazheniiakh-dlia/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1277-kadrovyi-kuratorskii-freimvork-s-uchetom-neopredelennosti/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1278-ispolzovanie-bazy-delta-dlia-peresmotra-tselei-uluchshaet-prognozirovanie-na/2026-06-17weekly0.7 https://korshunov.ai/en/article/1278-delta-based-target-reformulation-improves-electricity-load-forecasting/2026-06-17weekly0.7 https://korshunov.ai/en/article/1277-confusion-aware-transfer-teacher-curriculum-learning-framework/2026-06-17weekly0.7 https://korshunov.ai/en/article/1276-vision-language-models-don-t-always-need-images-for-chest-x-ray-accuracy/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1274-nevidimoe-vosstanovlenie-skrytykh-domenov-za-schet-otkrytiia-simmetrii-bez/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1275-smaa-fair-rasshirenie-smaa-s-uchetom-spravedlivosti-dlia-ranzhirovaniia/2026-06-17weekly0.7 https://korshunov.ai/en/article/1274-blind-recovery-of-latent-domains-via-unsupervised-symmetry-discovery/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1273-legkie-eksperimentalnye-skrytye-vospominaniia-dlia-nepreryvnogo-samouluchsheniia/2026-06-17weekly0.7 https://korshunov.ai/en/article/1275-smaa-fair-a-fairness-aware-extension-of-smaa-for-ranking/2026-06-17weekly0.7 https://korshunov.ai/en/article/1273-lightweight-experiential-latent-memories-for-continual-self-improvement/2026-06-17weekly0.7 https://korshunov.ai/en/article/1270-conservation-laws-for-modern-neural-architectures/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1272-querymarket-onlain-aktivnoe-obuchenie-s-uchetom-stoimosti-v-rynkakh-dannykh/2026-06-17weekly0.7 https://korshunov.ai/en/article/1272-querymarket-cost-aware-online-active-learning-in-data-markets/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1270-zakony-sokhraneniia-dlia-sovremennykh-neironnykh-arkhitektur/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1271-net-besplatnoi-spravedlivosti-fundamentalnye-ogranicheniia-v-sistemakh/2026-06-17weekly0.7 https://korshunov.ai/en/article/1271-no-free-fairness-fundamental-limits-in-learning-systems/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1267-qwen-robotmanip-dostigaet-obobshcheniia-v-robotizirovannoi-manipuliatsii/2026-06-17weekly0.7 https://korshunov.ai/en/article/1267-qwen-robotmanip-achieves-generalization-in-robotic-manipulation/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1268-stabilizatsiia-ubezhdenii-bolshikh-iazykovykh-modelei-za-schet-promptovannogo/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1269-funktsionalnaia-ekvivalentnost-v-vnimanii-s-pozitsionnymi-kodami/2026-06-17weekly0.7 https://korshunov.ai/en/article/1269-functional-equivalence-in-attention-with-positional-encodings/2026-06-17weekly0.7 https://korshunov.ai/en/article/1268-llm-belief-stabilization-via-prompted-predictive-resampling/2026-06-17weekly0.7 https://korshunov.ai/en/article/1266-wallzero-beats-go-pros-in-wallgo/2026-06-17weekly0.7 https://korshunov.ai/en/article/1264-anchorkv-safety-aware-kv-cache-compression-with-refusal-anchor/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1264-anchorkv-bezopasnost-orientirovannoe-szhatie-kesha-kv-s-ispolzovaniem-markera/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1265-metaklassifikatsiia-odnoklassovykh-modelei-cherez-ranzhirovanie-i-blizhaishikh/2026-06-17weekly0.7 https://korshunov.ai/en/article/1265-meta-classification-of-one-class-models-via-ranking-and-nearest-neighbor/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1266-wallzero-pobezhdaet-professionalnykh-igrokov-go-v-wallgo/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1261-razmernost-opredeliaet-usloviia-pri-kotorykh-modulnost-pomogaet-v-nepreryvnom/2026-06-17weekly0.7 https://korshunov.ai/en/article/1262-mkan-monotonic-kolmogorov-arnold-networks-with-hard-monotonicity/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1262-mkan-monotonnye-kolmogorov-arnoldovye-seti-s-zhestkoi-monotonnostiu/2026-06-17weekly0.7 https://korshunov.ai/en/article/1263-order-independent-cell-level-representations-for-multi-task-table-recognition/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1263-poriadkovo-nezavisimye-predstavleniia-na-urovne-iacheek-dlia-mnogozadachnoi/2026-06-17weekly0.7 https://korshunov.ai/en/article/1261-dimensionality-controls-when-modularity-helps-in-continual-learning/2026-06-17weekly0.7 https://korshunov.ai/en/article/1258-softmoe-soft-differentiable-routing-for-mixture-of-experts-in-llms/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1258-softmoe-miagkaia-differentsiruemaia-marshrutizatsiia-dlia-smesi-ekspertov-v/2026-06-17weekly0.7 https://korshunov.ai/en/article/1260-kanlib-a-modular-and-efficient-kolmogorov-arnold-network-framework/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1260-kanlib-modulnaia-i-effektivnaia-sistema-kolmogorova-arnolda/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1259-gibridnaia-model-ret-dnn-s-xgboost-dlia-prognozirovaniia-povedeniia-klientov/2026-06-17weekly0.7 https://korshunov.ai/en/article/1259-hybrid-ret-dnn-with-xgboost-for-customer-behavior-forecasting/2026-06-17weekly0.7 https://korshunov.ai/en/article/1255-mcwc-forecasting-with-cyclicity-trend-and-channel-correlation/2026-06-17weekly0.7 https://korshunov.ai/en/article/1256-differential-privacy-in-gaussian-process-posterior-sampling/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1256-differentsialnaia-privatnost-v-vyborke-posteriornogo-raspredeleniia-gaussovogo/2026-06-17weekly0.7 https://korshunov.ai/en/article/1257-cers-cot-enhanced-reasoning-for-medical-image-segmentation/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1257-cers-uluchshennoe-rassuzhdenie-na-osnove-tsepochki-myslei-dlia-segmentatsii/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1255-mcwc-prognozirovanie-s-tsiklichnostiu-tendentsiei-i-korreliatsiei-kanalov/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1254-polu-sviaz-v-dostatochnosti-znanii-grafov-fundamentalnykh-modelei/2026-06-17weekly0.7 https://korshunov.ai/en/article/1252-blitz-fast-and-calibrated-nonparametric-conditional-independence-test/2026-06-17weekly0.7 https://korshunov.ai/en/article/1254-half-link-sufficiency-in-knowledge-graph-foundation-models/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1252-blitz-bystryi-i-kalibrovannyi-neparametricheskii-test-nezavisimosti-uslovii/2026-06-17weekly0.7 https://korshunov.ai/en/article/1253-c2fl-clustered-continual-federated-learning-under-spatial-and-temporal-drift/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1253-c2fl-klasternoe-nepreryvnoe-federativnoe-obuchenie-pri-prostranstvennoi-i-dreife/2026-06-17weekly0.7 https://korshunov.ai/en/article/1251-star-spatiotemporal-adaptive-reward-allocation-for-text-to-image-rl-post/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1251-star-spatiotemporal-adaptive-reward-allocation-dlia-generatsii-teksta-v-v-rl/2026-06-17weekly0.7 https://korshunov.ai/en/article/1250-semantics-first-latent-modeling-for-3d-mri-reconstruction/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1249-mcwc-prognozirovanie-s-tsiklichnostiu-tendentsiei-i-korreliatsiei-kanalov/2026-06-17weekly0.7 https://korshunov.ai/en/article/1249-mcwc-forecasting-with-cyclicity-trend-and-channel-correlation/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1250-semanticheskii-podkhod-k-modelirovaniiu-skrytogo-prostranstva-dlia-3d-mrt/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1247-c2fl-klasternoe-nepreryvnoe-federativnoe-obuchenie-pri-prostranstvennoi-i-dreife/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1248-t-api-sovmestimyi-tsikl-react-dlia-opticheskikh-setei/2026-06-17weekly0.7 https://korshunov.ai/en/article/1248-t-api-compliant-react-loop-for-optical-networks/2026-06-17weekly0.7 https://korshunov.ai/en/article/1247-c2fl-clustered-continual-federated-learning-under-spatial-and-temporal-drift/2026-06-17weekly0.7 https://korshunov.ai/en/article/1246-llm-consumer-behavior-theory-a-new-research-field/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1246-teoriia-povedeniia-potrebitelei-na-osnove-bolshikh-iazykovykh-modelei-novaia/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1244-loopcoder-v2-dostigaet-optimalnoi-proizvoditelnosti-pri-dvukh-tsiklakh/2026-06-17weekly0.7 https://korshunov.ai/en/article/1244-loopcoder-v2-achieves-optimal-two-loop-performance/2026-06-17weekly0.7 https://korshunov.ai/en/article/1245-legalhallulens-auditing-hallucinations-in-legal-ai/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1245-legalhallulens-audit-galliutsinatsii-v-pravovom-ii/2026-06-17weekly0.7 https://korshunov.ai/en/article/1242-source-language-effects-in-cross-lingual-in-context-learning/2026-06-17weekly0.7 https://korshunov.ai/en/article/1241-provenanceguard-source-aware-factuality-verification-for-mcp-based-llm-agents/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1242-vliianie-iskhodnogo-iazyka-v-mezhiazykovom-obuchenii-v-kontekste/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1241-provenanceguard-proverka-faktichnosti-s-uchetom-istochnika-dlia-agentov-na-llm-s/2026-06-17weekly0.7 https://korshunov.ai/en/article/1243-catastrophic-forgetting-is-low-rank-a-function-space-theory/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1243-katastroficheskoe-zabvenie-eto-nizkoe-rangovoe-teoriia-funktsionalnogo/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1238-bezopasnost-i-privatnost-v-dialogakh-polzovatelei-s-llm/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1240-sinteticheskii-lichnyi-opyt-ii-v-podderzhke-ukhodchika/2026-06-17weekly0.7 https://korshunov.ai/en/article/1238-security-and-privacy-prompts-in-user-llm-conversations/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1239-pseudobench-otsenka-ustoichivosti-agentnykh-sistem-avtonomnogo-poiska-k/2026-06-17weekly0.7 https://korshunov.ai/en/article/1239-pseudobench-benchmarking-agentic-auto-research-resistance-to-pseudoscience/2026-06-17weekly0.7 https://korshunov.ai/en/article/1240-ai-s-synthetic-lived-experience-in-caregiver-support/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1236-kadrovaia-ii-arkhitektura-snizhaet-oshibki-diagnostiki-v-zdravookhranenii/2026-06-17weekly0.7 https://korshunov.ai/en/article/1237-scafe-using-llms-to-extract-clinically-meaningful-scar-features/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1237-scafe-ispolzovanie-bolshikh-iazykovykh-modelei-dlia-izvlecheniia-klinicheski/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1235-modeli-volterry-vvodiat-drobnuiu-shum-dlia-generatsii-na-osnove-scheta/2026-06-17weekly0.7 https://korshunov.ai/en/article/1235-volterra-generative-models-introduce-fractional-noise-for-score-based-generation/2026-06-17weekly0.7 https://korshunov.ai/en/article/1236-agentic-ai-framework-reduces-diagnostic-errors-in-healthcare/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1233-eagg-generatsiia-zakhvata-s-uchetom-korpusa-cherez-geometricheski-osoznannye/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1232-s4op-metod-priamogo-uproshcheniia-na-urovne-operatora-dlia-effektivnoi-ssm/2026-06-17weekly0.7 https://korshunov.ai/en/article/1233-eagg-embodiment-aligned-grasp-generation-via-geometry-aware-graph-conditioning/2026-06-17weekly0.7 https://korshunov.ai/en/article/1232-s4op-operator-level-pruning-for-efficient-ssm-deployment/2026-06-17weekly0.7 https://korshunov.ai/en/article/1234-hygrag-unified-framework-for-context-and-relation-aware-graph-rag/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1234-hygrag-edinaia-platforma-dlia-kontekst-i-otnoshenie-orientirovannykh-grafovykh/2026-06-17weekly0.7 https://korshunov.ai/en/article/1229-alerce-launches-text-to-sql-system-with-llms/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1231-isabellm-ai-driven-theorem-proving-for-consensus-verification/2026-06-17weekly0.7 https://korshunov.ai/en/article/1231-isabellm-ai-driven-theorem-proving-for-consensus-verification/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1229-alerce-zapuskaet-sistemu-tekst-v-sul-s-ispolzovaniem-bolshikh-iazykovykh-modelei/2026-06-17weekly0.7 https://korshunov.ai/en/article/1230-quality-aware-self-distillation-for-gui-grounding/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1230-kachestvosoznatelnaia-samodistilliatsiia-dlia-podkliucheniia-interfeisa/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1226-pervoe-dokazatelstvo-vtoraia-partiia-ii-proveren-na-zadachakh-issledovatelskogo/2026-06-17weekly0.7 https://korshunov.ai/en/article/1226-first-proof-second-batch-ai-tested-on-research-level-math-problems/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1228-izuchenie-spravedlivykh-pareto-optimalnykh-politik-v-mnogokriterialnoi-robastnoi/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1227-ternarnyi-mamba-predobuchennaia-qat-dlia-effektivnogo-szhatiia-ssm/2026-06-17weekly0.7 https://korshunov.ai/en/article/1227-ternary-mamba-pretrained-qat-for-efficient-ssm-compression/2026-06-17weekly0.7 https://korshunov.ai/en/article/1228-learning-fair-pareto-optimal-policies-in-multi-objective-reinforcement-learning/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1224-vstroennaia-rabochaia-sreda-ml-dlia-periferiinykh-ustroistv-na/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1225-uiazvimost-vvoda-s-pomoshchiu-troinykh-figurnykh-skobok-v-handlebars-vnedriat/2026-06-17weekly0.7 https://korshunov.ai/en/article/1225-handlebars-triple-brace-injection-exploits-structural-role-delimiters/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1223-vvedenie-v-cognitive-atrosphy-bench-dlia-vzaimodeistvii-llm-v-oblasti-zdorovia/2026-06-17weekly0.7 https://korshunov.ai/en/article/1223-introducing-cognitive-atrophy-bench-for-llm-mental-health-interactions/2026-06-17weekly0.7 https://korshunov.ai/en/article/1224-embedded-ml-workflow-for-microcontroller-edge-devices/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1222-pereispolzovanie-metaknowledge-v-obuchenii-s-usileniem/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1220-tac-pervyi-benchmark-agentov-po-voprosam-blagopoluchiia-zhivotnykh-v-ii/2026-06-17weekly0.7 https://korshunov.ai/en/article/1220-tac-first-agentic-benchmark-for-animal-welfare-in-ai/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1221-vvedenie-c3gd-publichnyi-nabor-audio-dannykh-o-vystrelakh/2026-06-17weekly0.7 https://korshunov.ai/en/article/1221-introducing-c3gd-a-public-gunshot-audio-dataset/2026-06-17weekly0.7 https://korshunov.ai/en/article/1222-meta-knowledge-reutilization-in-reinforcement-learning/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1219-prodolzhitelnost-raboty-flesh-pamiati-kak-iznosa-kapitala-v-pamiati-robota/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1218-weqa-voprosy-o-zdorove-v-nosimykh-ustroistvakh-s-adaptivnym-zaprosom-agentskoi/2026-06-17weekly0.7 https://korshunov.ai/en/article/1215-measurement-gap-in-eu-law-automation/2026-06-17weekly0.7 https://korshunov.ai/en/article/1219-flash-endurance-as-depreciating-capital-in-robot-memory/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1215-razryv-v-izmerenii-v-zakonodatelstve-evropeiskogo-soiuza-po-avtomatizatsii/2026-06-17weekly0.7 https://korshunov.ai/en/article/1218-weqa-wearable-health-question-answering-with-query-adaptive-agentic-reasoning/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1217-leads-agentskoe-otkrytie-gibridnykh-modelei-dlia-kardialnoi-elektrofiziologii/2026-06-17weekly0.7 https://korshunov.ai/en/article/1216-reage3d-realistic-3d-face-re-aging-with-view-consistency/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1216-reage3d-realistichnoe-3d-preobrazovanie-vozrasta-lits-s-sokhraneniem-vidov/2026-06-17weekly0.7 https://korshunov.ai/en/article/1217-leads-agentic-discovery-of-hybrid-models-for-cardiac-electrophysiology/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1212-kolmogorovskaia-regressiia-dlia-ustoichivykh-diffuzionnykh-politik/2026-06-17weekly0.7 https://korshunov.ai/en/article/1213-iuu-db-llm-driven-database-for-illegal-fishing-and-supply-chain-crimes/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1213-iuu-db-sistema-na-osnove-bolshikh-iazykovykh-modelei-dlia-otslezhivaniia-i-i-v/2026-06-17weekly0.7 https://korshunov.ai/en/article/1214-oracle-signals-in-agent-authored-test-code/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1214-signaly-orakla-v-kode-testov-napisannom-agentami/2026-06-17weekly0.7 https://korshunov.ai/en/article/1212-kolmogorov-regression-for-robust-diffusion-policies/2026-06-17weekly0.7 https://korshunov.ai/en/article/1211-drflow-benchmark-for-personalized-workflow-prediction/2026-06-17weekly0.7 https://korshunov.ai/en/article/1209-red-team-study-finds-frontier-llms-remain-vulnerable-to-adaptive-attacks/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1209-issledovanie-gruppy-krasnykh-komand-pokazyvaet-chto-peredovye-modeli-llm-pered/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1211-drflow-benchmark-dlia-predskazaniia-personalizirovannykh-rabochikh-protsessov/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1210-vneshnii-vypusk-dannykh-po-zaiavkam-stenforda/2026-06-17weekly0.7 https://korshunov.ai/en/article/1210-stanford-edgar-filings-dataset-released/2026-06-17weekly0.7 https://korshunov.ai/en/article/1208-rubricstree-scalable-evaluation-framework-for-personal-health-agents/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1207-fprm-model-s-adaptivnym-vychisleniem-na-osnove-fiksirovannoi-tochki/2026-06-17weekly0.7 https://korshunov.ai/en/article/1207-fprm-fixed-point-reasoning-model-with-adaptive-compute/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1206-modeli-tsiklicheskogo-mira-obespechivaiut-effektivnost-na-100x/2026-06-17weekly0.7 https://korshunov.ai/en/article/1206-looped-world-models-achieve-100x-parameter-efficiency/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1208-rubricstree-masshtabiruemaia-sistema-otsenki-dlia-personalnykh-agentov/2026-06-17weekly0.7 https://korshunov.ai/en/article/1205-learning-red-agent-policy-from-observations-for-neurosymbolic-cyber-agents/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1205-obuchenie-politiki-krasnogo-agenta-na-osnove-nabliudenii-dlia-kiberagentov/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1203-reprorepo-masshtabirovanie-auditov-vosproizvodimosti-s-pomoshchiu-issue-github/2026-06-17weekly0.7 https://korshunov.ai/en/article/1203-reprorepo-scaling-reproducibility-audits-with-github-issues/2026-06-17weekly0.7 https://korshunov.ai/en/article/1204-evolvenav-self-evolving-memory-for-zero-shot-navigation/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1204-evolvenav-samorazvivaiushchaiasia-pamiat-dlia-navigatsii-bez-obucheniia/2026-06-17weekly0.7 https://korshunov.ai/en/article/1202-visual-verification-enables-inference-time-steering-and-autonomous-policy/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1200-otritsatelnaia-filtratsiia-tokenov-dlia-stabilnogo-odnorazovogo-obucheniia-rl/2026-06-17weekly0.7 https://korshunov.ai/en/article/1200-negative-token-filtering-for-stable-single-rollout-rl/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1202-vizualnaia-proverka-obespechivaet-upravlenie-na-etape-inferensa-i-avtonomnoe/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1197-mogut-li-iazykovye-modeli-obnaruzhit-nol/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1198-miagkoe-podskazyvanie-dlia-sobliudeniia-iazyka-v-multimodalnykh-llm/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1199-podacha-sistemy-odnovremennogo-perevoda-rechi-mllp-vrain-na-iwslt-2026/2026-06-17weekly0.7 https://korshunov.ai/en/article/1198-soft-prompting-for-language-adherence-in-multimodal-llms/2026-06-17weekly0.7 https://korshunov.ai/en/article/1199-mllp-vrain-s-simultaneous-speech-translation-submission-for-iwslt-2026/2026-06-17weekly0.7 https://korshunov.ai/en/article/1197-can-language-models-discover-zero/2026-06-17weekly0.7 https://korshunov.ai/en/article/1194-llm-generated-stories-show-low-diversity/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1195-speechdx-mnogotselevoi-benchmark-dlia-klinicheskoi-rechevoi-ii/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1194-istorii-sgenerirovannye-llm-pokazyvaiut-nizkuiu-raznoobrazie/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1196-proizvoditelnost-word2vec-v-minimalnoi-leksike-toki-pona/2026-06-17weekly0.7 https://korshunov.ai/en/article/1195-speechdx-multi-task-benchmark-for-clinical-speech-ai/2026-06-17weekly0.7 https://korshunov.ai/en/article/1196-word2vec-s-performance-in-toki-pona-s-minimal-vocabulary/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1192-iavnoe-protiv-skrytogo-podskazki-v-lvlms-dlia-otsylki-k-kommunikatsii/2026-06-17weekly0.7 https://korshunov.ai/en/article/1193-operationalizing-ontology-for-untranslatability-in-nlp/2026-06-17weekly0.7 https://korshunov.ai/en/article/1191-visuals-lie-consistency-speaks-disentangling-spatial-attention-from-reliability/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1191-vizualnye-dannye-lgut-soglasovannost-govorit-razedinenie-prostranstvennoi-ot-v/2026-06-17weekly0.7 https://korshunov.ai/en/article/1192-implicit-vs-explicit-prompting-in-lvlms-for-referential-communication/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1193-operatsionizatsiia-ontologii-dlia-neperevodimykh-sluchaev-v-nlp/2026-06-17weekly0.7 https://korshunov.ai/en/article/1190-narrativeworldbench-and-n-vssm-for-long-horizon-audio-drama/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1190-narrativeworldbench-i-n-vssm-dlia-dlitelnykh-audiodramaticheskikh-stsenariev/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1189-smeshchenie-rekomendatsii-llm-i-dinamika-konkurentsii-brendov/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1188-mode-rag-otsenka-i-sokrashchenie-khalliutsinatsii-v-m-rag/2026-06-17weekly0.7 https://korshunov.ai/en/article/1188-mode-rag-evaluating-and-reducing-hallucinations-in-m-rag/2026-06-17weekly0.7 https://korshunov.ai/en/article/1189-llm-recommendation-bias-and-brand-competition-dynamics/2026-06-17weekly0.7 https://korshunov.ai/en/article/1187-parse-real-document-defense-for-llm-agents/2026-06-17weekly0.7 https://korshunov.ai/en/article/1185-statewitness-activation-explainer-for-deception-auditing-in-llms/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1185-statewitness-obiasnitel-aktivatsii-dlia-audita-lzhi-v-llms/2026-06-17weekly0.7 https://korshunov.ai/en/article/1186-aipatient-arena-ehr-grounded-evaluation-of-llms-in-clinical-workflows/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1186-aipatient-arena-otsenka-llm-v-klinicheskikh-rabochikh-protsessakh-na-osnove-erv/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1187-parse-zashchita-realnykh-dokumentov-dlia-agentov-na-osnove-llm/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1180-llms-prevoskhodiat-liudei-po-predskazaniiu-sleduiushchego-govoriashchego/2026-06-17weekly0.7 https://korshunov.ai/en/article/1184-second-order-bias-in-llms-evaluating-judgment-based-bias/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1184-vtoroi-poriadok-smeshcheniia-v-llm-otsenka-smeshcheniia-na-osnove-suzhdenii/2026-06-17weekly0.7 https://korshunov.ai/en/article/1183-routing-accuracy-degradation-and-recovery-in-enterprise-agent-systems/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1183-padenie-i-vosstanovlenie-tochnosti-marshrutizatsii-v-sistemakh-agentov/2026-06-17weekly0.7 https://korshunov.ai/en/article/1180-llms-outperform-humans-in-next-speaker-prediction/2026-06-17weekly0.7 https://korshunov.ai/en/article/1182-expressivity-analysis-of-hierarchical-modelling-in-deep-transformers/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1182-analiz-vyrazitelnosti-ierarkhicheskikh-modelei-v-glubokikh-transformerakh/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1181-dekodirovanie-nar-mbr-dlia-bystroi-i-tochnoi-raspoznavaniia-rechi/2026-06-17weekly0.7 https://korshunov.ai/en/article/1181-nar-mbr-decoding-for-fast-and-accurate-speech-recognition/2026-06-17weekly0.7 https://korshunov.ai/en/article/1177-opd-evolver-on-policy-distillation-for-holistic-agent-evolving/2026-06-17weekly0.7 https://korshunov.ai/en/article/1178-pruned-llms-fail-in-open-generation-despite-passing-multiple-choice/2026-06-17weekly0.7 https://korshunov.ai/en/article/1179-llm-features-can-hurt-gnns-via-concatenation-interference/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1178-szhatye-modeli-iazykovykh-modelei-ne-spravliaiutsia-s-otkrytym-generirovaniem-na/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1179-funktsii-llm-mogut-navredit-gnn-cherez-interferentsiiu-pri-konkatenatsii/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1177-opd-evolver-on-politicheskaia-distilliatsiia-dlia-vsestoronnego-agentov/2026-06-17weekly0.7 https://korshunov.ai/en/article/1176-prompt-perturbation-for-reliable-llm-evaluation/2026-06-17weekly0.7 https://korshunov.ai/en/article/1175-skillmigrator-enables-cross-site-web-skill-transfer-via-layout-matching/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1175-skillmigrator-obespechivaet-peredachu-navykov-veb-saitov-cherez-sovpadenie/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1174-mambacount-effektivnyi-tekst-orientirovannyi-podschet-obektov/2026-06-17weekly0.7 https://korshunov.ai/en/article/1174-mambacount-efficient-text-guided-object-counting/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1176-perturbation-zaprosa-dlia-nadezhnoi-otsenki-bolshikh-iazykovykh-modelei/2026-06-17weekly0.7 https://korshunov.ai/en/article/1172-llm-designed-training-environment-for-rl-with-multi-agent-reasoning/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1173-envrl-ispolzovanie-dinamiki-sredy-v-agentnom-obuchenii-s-pomoshchiu/2026-06-17weekly0.7 https://korshunov.ai/en/article/1173-envrl-leveraging-environment-dynamics-in-agentic-rl/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1172-obuchenie-llm-dlia-sredy-obucheniia-rl-s-mnogomodulnym-rassuzhdeniem/2026-06-17weekly0.7 https://korshunov.ai/en/article/1171-swifttrans-improves-llm-code-translation-efficiency/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1171-swifttrans-povyshaet-effektivnost-perevoda-koda-na-osnove-llm/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1167-modeli-vizualno-iazykovykh-sistem-ne-vsegda-trebuiut-izobrazhenii-dlia-tochnosti/2026-06-17weekly0.7 https://korshunov.ai/en/article/1169-llms-infer-cultural-context-but-fail-to-apply-it/2026-06-17weekly0.7 https://korshunov.ai/en/article/1170-suco-sufficiency-guided-continuous-adaptive-reasoning/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1169-llms-opredeliaiut-kulturnyi-kontekst-no-ne-primeniaiut-ego/2026-06-17weekly0.7 https://korshunov.ai/en/article/1167-vision-language-models-don-t-always-need-images-for-chest-x-ray-accuracy/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1170-suco-adaptivnoe-obosnovannoe-rassuzhdenie-osnovannoe-na-dostatochnosti/2026-06-17weekly0.7 https://korshunov.ai/en/article/1168-ecomagentbench-benchmarking-shopping-agents-with-hidden-intent/2026-06-17weekly0.7 https://korshunov.ai/en/article/1166-ai-driven-avatars-enable-realistic-act-psychotherapy-training/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1166-avtorskie-avatary-na-osnove-ii-obespechivaiut-realistichnuiu-podgotovku-k-po-akt/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1168-ecomagentbench-otsenka-agentov-pokupok-s-skrytoi-tseliu/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1165-antiparadoks-slopa-perepiska-ii-snizhaet-klinicheskuiu-neopredelennost-i/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1163-dife-proveriaet-proniknovenie-bekdora-v-clip-na-raznykh-interfeisakh/2026-06-17weekly0.7 https://korshunov.ai/en/article/1165-the-slop-paradox-ai-rewriting-degrades-clinical-uncertainty-and-cross-modal/2026-06-17weekly0.7 https://korshunov.ai/en/article/1163-dife-audits-clip-backdoor-exposure-across-deployment-interfaces/2026-06-17weekly0.7 https://korshunov.ai/en/article/1164-coding-benchmarks-misaligned-with-agentic-software-engineering/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1164-nesootvetstvie-kodovykh-standartov-agentnoi-razrabotke-programmnogo/2026-06-17weekly0.7 https://korshunov.ai/en/article/1161-bilingual-fine-tuning-improves-low-resource-asr-with-language-identification/2026-06-17weekly0.7 https://korshunov.ai/en/article/1162-a-framework-for-evaluating-agentic-skills-at-scale/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1160-multiclin-benchmark-dlia-multiskriptovoi-asr-v-klinicheskikh-usloviiakh/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1162-kontseptualnaia-ramka-otsenki-agentnykh-navykov-na-masshtabe/2026-06-17weekly0.7 https://korshunov.ai/en/article/1160-multiclin-benchmark-for-multiscript-asr-in-clinical-settings/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1161-dvukhiazychnaia-nastroika-uluchshaet-avtomaticheskoe-raspoznavanie-rechi-v-s-s/2026-06-17weekly0.7 https://korshunov.ai/en/article/1159-self-supervised-speech-models-lack-tonal-context-compensation/2026-06-17weekly0.7 https://korshunov.ai/en/article/1158-automated-prompt-optimization-for-llm-game-agents/2026-06-17weekly0.7 https://korshunov.ai/en/article/1157-gamecraft-bench-evaluating-end-to-end-game-generation/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1157-gamecraft-bench-otsenka-polnogo-generirovaniia-igr/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1158-avtomatizirovannaia-optimizatsiia-promptov-dlia-agentov-na-osnove-llm/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1159-samoobuchaemye-modeli-rechi-ne-uchityvaiut-kompensatsiiu-tonalnogo-konteksta/2026-06-17weekly0.7 https://korshunov.ai/en/article/1154-non-negative-elastic-net-decoding-for-information-retrieval/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1155-chlogic-proverka-ustoichivosti-logicheskogo-myshleniia-v-kitaiskikh/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1154-nepolozhitelnoe-dekodirovanie-elastichnogo-seti-dlia-informatsionnogo-poiska/2026-06-17weekly0.7 https://korshunov.ai/en/article/1155-chlogic-testing-logical-reasoning-robustness-in-chinese-expressions/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1156-dinamicheskoe-redaktirovanie-posledovatelnosti-umenshaet-peregruzku-myshleniia-v/2026-06-17weekly0.7 https://korshunov.ai/en/article/1156-dynamic-rollout-editing-reduces-overthinking-in-rl-trained-reasoning-models/2026-06-17weekly0.7 https://korshunov.ai/en/article/1151-voidpadding-decoupling-eos-termination-and-padding-in-mdlms/2026-06-17weekly0.7 https://korshunov.ai/en/article/1152-fine-tuning-llms-for-passive-depression-severity-estimation/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1151-voidpadding-razdelenie-eos-zaversheniia-i-zapolneniia-v-mdlms/2026-06-17weekly0.7 https://korshunov.ai/en/article/1153-interventional-post-training-of-speech-foundation-models/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1153-interventsionnoe-postobuchivanie-rechevykh-fundamentalnykh-modelei/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1152-nastroika-llm-dlia-otsenki-stepeni-passivnoi-depressii/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1150-ozhidanie-v-modeliakh-rassuzhdenii-vyzyvaet-u-menia-drozh/2026-06-17weekly0.7 https://korshunov.ai/en/article/1150-wait-in-reasoning-models-makes-my-eye-twitch/2026-06-17weekly0.7 https://korshunov.ai/en/article/1149-someone-awhile-ago-did-a-quant-shootout-for-qwen3-6/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1149-kto-to-ne-tak-davno-provel-sorevnovanie-po-kvantovaniiu-dlia-qwen3-6/2026-06-17weekly0.7 https://korshunov.ai/en/article/1148-vibethinker-3b-what-is-this-witchcraft/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1148-vibethinker-3b-chto-za-koldovstvo/2026-06-17weekly0.7 https://korshunov.ai/ru/article/1146-datasette-tailscale-0-1a0-alpha-release/2026-06-17weekly0.7 https://korshunov.ai/en/article/1146-datasette-tailscale-0-1a0-alpha-release/2026-06-17weekly0.7 https://korshunov.ai/en/article/1145-quantitative-comparison-of-qwen3-6-model-performance/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1145-kolichestvennyi-analiz-proizvoditelnosti-modeli-qwen3-6/2026-06-16weekly0.7 https://korshunov.ai/en/article/1142-i-didn-t-know-it-was-possible-to-compile-llamacpp-to-run-cuda-vulkan-at-the-same/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1142-ia-ne-znal-chto-vozmozhno-skompilirovat-llamacpp-dlia-odnovremennoi-raboty-s-i/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1141-le-gros-chaton-rabotaet-na-radio-84-korolla/2026-06-16weekly0.7 https://korshunov.ai/en/article/1141-le-gros-chaton-running-on-my-84-corolla-radio/2026-06-16weekly0.7 https://korshunov.ai/en/article/1138-minimax-m3-4-bit-mlx-initial-benchmark-on-mac-studio-m3-with-512gb/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1138-minimax-m3-4-bit-mlx-pervonachalnyi-benchmark-na-mac-studio-m3-s-512-gb/2026-06-16weekly0.7 https://korshunov.ai/en/article/1136-glm-5-2-now-first-on-design-arena/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1136-glm-5-2-teper-pervyi-na-design-arena/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1134-iavliaetsia-li-le-gros-chaton-otkrytoi-sistemoi/2026-06-16weekly0.7 https://korshunov.ai/en/article/1134-is-le-gros-chaton-open-source/2026-06-16weekly0.7 https://korshunov.ai/en/article/1132-glm-5-2-releases-open-weights-with-strong-coding-performance/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1133-zhipu-vyros-na-33-posle-togo-kak-birzhevye-rynki-uvelichili-stavki-na-kitaiskuiu/2026-06-16weekly0.7 https://korshunov.ai/en/article/1133-zhipu-surges-33-as-wall-street-raises-bets-on-china-ai-after-anthropic-curbs/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1132-glm-5-2-vypustil-otkrytye-vesa-s-silnoi-proizvoditelnostiu-v-programmirovanii/2026-06-16weekly0.7 https://korshunov.ai/en/article/1126-llama-cpp-releases-b9660-with-fixes-and-new-binaries/2026-06-16weekly0.7 https://korshunov.ai/en/article/1127-v2-1-179-release-notes/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1127-primechaniia-k-vypusku-v2-1-179/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1126-llama-cpp-vypuskaet-b9660-s-ispravleniiami-i-novymi-binarnymi-failami/2026-06-16weekly0.7 https://korshunov.ai/en/article/1123-evalatro-an-open-benchmark-where-llms-play-real-balatro/2026-06-16weekly0.7 https://korshunov.ai/en/article/1125-glm-5-2-api-live-weights-on-hugging-face-ollama-support/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1125-zhivoi-api-glm-5-2-vesa-na-hugging-face-podderzhka-ollama/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1124-my-otkryli-iskhodnyi-kod-nashego-agenta-na-osnove-bolshikh-iazykovykh-modelei/2026-06-16weekly0.7 https://korshunov.ai/en/article/1124-we-open-sourced-our-llm-based-qa-agent-to-catch-breakages-faster/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1123-evalatro-otkrytyi-benchmark-gde-llmy-igraiut-realnuiu-balatro/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1122-samyi-deshevyi-zhelezo-dlia-qwen-3-6-modeli-27b-i-35b-a3b/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1121-ostanovites-ispolzovat-ollama/2026-06-16weekly0.7 https://korshunov.ai/en/article/1121-stop-using-ollama/2026-06-16weekly0.7 https://korshunov.ai/en/article/1122-cheapest-hardware-for-qwen-3-6-27b-and-35b-a3b-models/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1119-langgraph-cli-0-4-30-vypushchen/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1120-osnovatel-hashicorp-utverzhdaet-chto-lokalnye-modeli-poka-ne-dostatochno/2026-06-16weekly0.7 https://korshunov.ai/en/article/1120-hashicorp-founder-claims-local-models-aren-t-good-enough-yet/2026-06-16weekly0.7 https://korshunov.ai/en/article/1119-langgraph-cli-0-4-30-released/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1118-primechaniia-po-vypusku-claude-v2-1-178/2026-06-16weekly0.7 https://korshunov.ai/en/article/1118-claude-v2-1-178-release-notes/2026-06-16weekly0.7 https://korshunov.ai/en/article/1114-llama-cpp-releases-version-b9672-with-boringssl-update/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1114-llama-cpp-vypustil-versiiu-b9672-s-obnovleniem-boringssl/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1107-tred-o-sozdanii-modeli-soobshchestva-vozmozhna-kraudsorsirovannaia-trenirovka/2026-06-16weekly0.7 https://korshunov.ai/en/article/1106-glm-5-2-crosses-80-on-terminal-bench/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1106-glm-5-2-prevyshaet-80-na-terminal-bench/2026-06-16weekly0.7 https://korshunov.ai/en/article/1108-benchmark-for-tiny-llms-on-natural-language-file-search/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1108-otsenka-nebolshikh-modelei-llm-na-poiske-failov-na-estestvennom-iazyke/2026-06-16weekly0.7 https://korshunov.ai/en/article/1107-community-model-build-thread-crowdsourced-training-feasible/2026-06-16weekly0.7 https://korshunov.ai/en/article/1105-release-branch-created-for-v1-38-0/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1105-sozdan-razvetvlennyi-kanal-dlia-versii-1-38-0/2026-06-16weekly0.7 https://korshunov.ai/en/article/1104-fable-and-mythos-model-welfare-analysis/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1104-analiz-blagopoluchiia-modeli-fable-i-mythos/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1103-glm-5-2-zanimaet-vtoroe-mesto-v-webdew-arena/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1102-glm-5-2-teper-dostupen-na-huggingchat/2026-06-16weekly0.7 https://korshunov.ai/en/article/1102-glm-5-2-now-available-on-huggingchat/2026-06-16weekly0.7 https://korshunov.ai/en/article/1103-glm-5-2-takes-2-spot-on-webdew-arena/2026-06-16weekly0.7 https://korshunov.ai/en/article/1100-mistral-announces-new-family-of-open-weight-models-in-july/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1101-glimmer-1-fundamentalnaia-model-iazyka-s-10-000-parametrami/2026-06-16weekly0.7 https://korshunov.ai/en/article/1101-glimmer-1-a-10000-parameter-foundational-language-model/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1100-mistral-obiavliaet-o-novoi-seme-otkrytykh-modelei-v-iiule/2026-06-16weekly0.7 https://korshunov.ai/en/article/1099-zai-org-releases-glm-5-2/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1099-zai-org-opublikoval-glm-5-2/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1093-bartowski-command-a-plus-05-2026-gguf-na-hugging-face/2026-06-16weekly0.7 https://korshunov.ai/en/article/1093-bartowski-command-a-plus-05-2026-gguf-on-hugging-face/2026-06-16weekly0.7 https://korshunov.ai/en/article/1089-anyone-running-qwen-3-6-27b-ud-q8-on-multiple-gpus/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1089-kto-nibud-zapuskaet-qwen-3-6-27b-ud-q8-na-neskolkikh-gpu/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1079-georgi-gerganov-khvalit-qwen3-6-27b-za-zadachi-po-programmirovaniiu/2026-06-16weekly0.7 https://korshunov.ai/en/article/1079-georgi-gerganov-praises-qwen3-6-27b-for-coding-tasks/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1078-luchshaia-model-i-nastroika-dlia-macbook-pro-s-128-gb-ozu-i-8-tb-m5-max/2026-06-16weekly0.7 https://korshunov.ai/en/article/1078-best-model-and-configuration-for-128gb-ram-8tb-m5-max-macbook-pro/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1077-prizyv-k-otkrytym-modeliam-i-prichiny-po-kotorym-my-ne-mozhem-doveriat-frontier/2026-06-16weekly0.7 https://korshunov.ai/en/article/1077-the-case-for-open-weight-models-and-why-we-can-t-trust-frontier-labs/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1072-issleduetsia-li-diffusiongemma-deistvitelno-nastolko-khorosh-v-agente-pi/2026-06-16weekly0.7 https://korshunov.ai/en/article/1072-is-diffusiongemma-really-that-good-in-a-pi-agent/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1069-anthropic-izmeniaet-pozitsiiu-po-ispolzovaniiu-claude-p-storonnimi-uchastnikami/2026-06-16weekly0.7 https://korshunov.ai/en/article/1069-anthropic-reversing-stance-on-claude-p-third-party-usage/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1066-vibethinker-3b-dostigaet-peredovykh-pokazatelei-po-matematike-i/2026-06-16weekly0.7 https://korshunov.ai/en/article/1066-vibethinker-3b-achieves-frontier-math-and-coding-performance/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1065-obiavlen-nabor-robotizirovannykh-instrumentov-qwen/2026-06-16weekly0.7 https://korshunov.ai/en/article/1065-qwen-robot-suite-announced/2026-06-16weekly0.7 https://korshunov.ai/en/article/1059-frontier-post-training-recipe-review-with-finbarr-timbers/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1059-obzor-retseptur-posle-trenirovki-v-frontier-s-finbarr-timbers/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1058-llama-cpp-vypuskaet-b96669-s-podderzhkoi-vybora-v-bekende-dlia-eagle3/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1057-llama-cpp-release-b9670-ispravleniia-i-novye-sborki/2026-06-16weekly0.7 https://korshunov.ai/en/article/1057-llama-cpp-release-b9670-fixes-and-new-builds/2026-06-16weekly0.7 https://korshunov.ai/en/article/1058-llama-cpp-releases-b96669-with-backend-sampling-for-eagle3/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1056-pochemu-diffusiongemma-mozhet-otlichno-spravliatsia-s-vyzovami-instrumentov-na/2026-06-16weekly0.7 https://korshunov.ai/en/article/1056-why-diffusiongemma-might-excel-at-tool-calls-despite-lower-base-quality/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1054-llama-cpp-release-b9667-dobavliaet-podderzhku-vulkan-i-cuda/2026-06-16weekly0.7 https://korshunov.ai/en/article/1054-llama-cpp-release-b9667-adds-vulkan-and-cuda-support/2026-06-16weekly0.7 https://korshunov.ai/en/article/1053-llama-cpp-release-b9668-adds-uma-host-visible-memory-and-cross-platform-binaries/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1052-rezultaty-testa-proizvoditelnosti-kvantovaniia-qwen3-6-27b/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1051-izuchenie-konteksta-i-inzhenerii-uderzhaniia-dlia-lokalnykh-ii/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1053-llama-cpp-release-b9668-dobavliaet-uma-host-visible-memory-i-binarniki-dlia/2026-06-16weekly0.7 https://korshunov.ai/en/article/1052-qwen3-6-27b-quantization-performance-test-results/2026-06-16weekly0.7 https://korshunov.ai/en/article/1051-learning-context-and-harness-engineering-for-local-first-ai/2026-06-16weekly0.7 https://korshunov.ai/en/article/1047-gemma-12b-reasoning-hardening-instructions/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1047-instruktsii-po-uprochneniiu-logicheskogo-myshleniia-dlia-gemma-12b/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1045-budte-ostorozhny-pered-ispolzovaniem-distillirovannykh-modelei-qwen-claude-oni/2026-06-16weekly0.7 https://korshunov.ai/en/article/1045-be-wary-of-qwen-claude-distillations-they-re-often-worse-than-the-base-model/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1044-llama-cpp-release-b9665-dobavliaet-flag-offline-i-novye-binarnye-sborki/2026-06-16weekly0.7 https://korshunov.ai/en/article/1044-llama-cpp-release-b9665-adds-offline-flag-and-new-binary-builds/2026-06-16weekly0.7 https://korshunov.ai/en/article/1041-donate-your-coding-sessions-to-an-open-cc-by-4-0-dataset/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1041-pozhertvuite-svoi-sessii-programmirovaniia-v-otkrytuiu-bazu-dannykh-pod-cc-by-4/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1038-aerollm-bystraia-otkrytaia-ii-prilozhenie-dlia-protsessorov-apple-silicon/2026-06-16weekly0.7 https://korshunov.ai/en/article/1038-aerollm-fast-open-source-ai-app-for-apple-silicon/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1039-sliianie-gpu-dlia-obucheniia-soobshchestvennoi-modeli/2026-06-16weekly0.7 https://korshunov.ai/en/article/1037-nex-n2-pro-is-the-real-deal/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1037-nex-n2-pro-eto-nastoiashchee-kachestvo/2026-06-16weekly0.7 https://korshunov.ai/en/article/1039-pooling-gpus-to-train-a-community-model/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1034-reliz-llama-cpp-b9663-dobavliaet-podderzhku-sycl-i-novye-binarnye-sborki/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1033-sycl-podderzhka-perestavlennykh-q4-k-q5-k-q6-k-moe-mul-mat-id/2026-06-16weekly0.7 https://korshunov.ai/en/article/1033-sycl-support-reordered-q4-k-q5-k-q6-k-moe-mul-mat-id/2026-06-16weekly0.7 https://korshunov.ai/en/article/1034-llama-cpp-release-b9663-adds-sycl-support-and-new-binary-builds/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1032-iavliaiutsia-li-malymi-lokalnymi-modeliami-dlia-avtomatizatsii-veshchi/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1031-kak-neispolzuiushchie-mac-polzovateli-zapuskaiut-deepseekv4-lokalno/2026-06-16weekly0.7 https://korshunov.ai/en/article/1032-are-small-local-models-for-automation-a-thing/2026-06-16weekly0.7 https://korshunov.ai/en/article/1031-how-are-non-mac-users-running-deepseekv4-locally/2026-06-16weekly0.7 https://korshunov.ai/en/article/1029-diffusion-gemma-jailbreak-allows-explicit-content/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1029-razryv-diffusion-gemma-pozvoliaet-generirovat-otkrovennyi-kontent/2026-06-16weekly0.7 https://korshunov.ai/en/article/1027-vulkan-adds-col2im-1d-op-and-supports-multiple-platforms/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1027-vulkan-dobavliaet-operatsiiu-col2im-1d-i-podderzhivaet-neskolko-platform/2026-06-16weekly0.7 https://korshunov.ai/en/article/1024-any-benefit-to-a-multi-machine-setup-for-local-llms/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1024-kakie-preimushchestva-predostavliaet-mnogomashinnaia-nastroika-dlia-lokalnykh/2026-06-16weekly0.7 https://korshunov.ai/en/article/1025-fable-5-export-controls-harm-us-cyber-defense/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1025-ogranicheniia-po-eksportu-fable-5-nanosiat-vred-zashchite-ssha-v-sfere/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1022-eshche-li-kvantovannye-modeli-generatsii-izobrazhenii-nakhodiatsia-v-stadii/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1021-nex2-mini-phase-twin-16-gb-model-30b-vypushchena/2026-06-16weekly0.7 https://korshunov.ai/en/article/1022-are-quantized-image-generation-models-still-wip/2026-06-16weekly0.7 https://korshunov.ai/en/article/1021-nex2-mini-phase-twin-16gb-footprint-30b-model-released/2026-06-16weekly0.7 https://korshunov.ai/en/article/1006-the-white-house-is-ratcheting-up-its-war-against-anthropic/2026-06-16weekly0.7 https://korshunov.ai/en/article/1020-dgx-spark-is-being-defamed/2026-06-16weekly0.7 https://korshunov.ai/en/article/1018-contrastive-difference-cka-reveals-concept-specific-alignment-across-llm/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1020-dgx-spark-podvergaetsia-nespravedlivoi-kritike/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1018-kontrastno-raznostnoe-cka-raskryvaet-kontseptualno-spetsificheskoe-vyravnivanie/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1006-belyi-dom-usilivaet-voinu-protiv-anthropic/2026-06-16weekly0.7 https://korshunov.ai/en/article/1019-symbolic-informalization-in-informath-project/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1019-simvolnaia-neformalizatsiia-v-proekte-informath/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1017-logos-obshchee-generativnoe-modelirovanie-dlia-estestvennykh-nauk/2026-06-16weekly0.7 https://korshunov.ai/en/article/1017-logos-a-general-purpose-generative-model-for-natural-sciences/2026-06-16weekly0.7 https://korshunov.ai/en/article/1016-less-is-more-adaptive-sampling-for-diffusion-language-models/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1016-menshe-bolshe-adaptivnaia-vyborka-dlia-modelei-diffuzionnogo-iazyka/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1015-reliz-dataseta-impacteen-v-angliiskoi-i-polskoi-versiiakh/2026-06-16weekly0.7 https://korshunov.ai/en/article/1015-impacteen-dataset-released-with-english-and-polish-versions/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1014-kliuchevye-svoistva-dlia-effektivnogo-rassuzhdeniia-kodovogo-interpretatora/2026-06-16weekly0.7 https://korshunov.ai/en/article/1014-key-properties-for-effective-code-interpreter-reasoning/2026-06-16weekly0.7 https://korshunov.ai/en/article/1012-tokenpilot-cache-efficient-context-management-for-llm-agents/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1011-deeprubric-effektivnaia-rl-dlia-agentov-glubokogo-issledovaniia/2026-06-16weekly0.7 https://korshunov.ai/en/article/1011-deeprubric-efficient-rl-for-deep-research-agents/2026-06-16weekly0.7 https://korshunov.ai/en/article/1013-post-hoc-operators-fail-to-improve-accuracy-in-small-code-models/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1013-posle-operatory-ne-uluchshaiut-tochnost-v-malykh-modeliakh-koda/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1012-tokenpilot-effektivnoe-upravlenie-kontekstom-dlia-agentov-llm/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1010-kveraser-effektivnoe-lokalizovannoe-udalenie-konteksta-v-llms/2026-06-16weekly0.7 https://korshunov.ai/en/article/1010-kveraser-efficient-localized-context-erasing-in-llms/2026-06-16weekly0.7 https://korshunov.ai/en/article/1009-metasyn-benchmarking-llm-agents-on-meta-analysis-articles/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1009-metasyn-otsenka-agentov-llm-na-statiakh-po-meta-analizu/2026-06-16weekly0.7 https://korshunov.ai/en/article/1008-contextrl-context-aware-rl-for-llms/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1008-contextrl-kontekstualno-optimizirovannyi-rl-dlia-llms/2026-06-16weekly0.7 https://korshunov.ai/en/article/966-symbolic-informalization-in-informath-project/2026-06-16weekly0.7 https://korshunov.ai/ru/article/965-semanticheskii-otrazhenie-sintez-obraztsov-za-predelami-raspredeleniia-dlia/2026-06-16weekly0.7 https://korshunov.ai/en/article/1007-language-models-encode-value-of-their-current-trajectory/2026-06-16weekly0.7 https://korshunov.ai/ru/article/966-simvolicheskaia-neformalizatsiia-v-proekte-informath/2026-06-16weekly0.7 https://korshunov.ai/en/article/965-semantic-flip-synthetic-ood-generation-for-robust-refusal/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1007-iazykovye-modeli-kodiruiut-znachenie-svoei-tekushchei-traektorii/2026-06-16weekly0.7 https://korshunov.ai/ru/article/963-opublikovano-mnozhestvo-dannykh-impacteen-v-angliiskoi-i-polskoi-versiiakh/2026-06-16weekly0.7 https://korshunov.ai/en/article/964-bintrack-open-source-spatial-qa-with-binary-trajectory-search/2026-06-16weekly0.7 https://korshunov.ai/ru/article/964-bintrack-otkrytyi-istochnik-spektralnyi-qa-s-dvoichnym-poiskom-traektorii/2026-06-16weekly0.7 https://korshunov.ai/en/article/963-impacteen-dataset-released-with-english-and-polish-versions/2026-06-16weekly0.7 https://korshunov.ai/ru/article/962-golod-uchitsia-zavisimost-ot-kanalov-voznagrazhdeniia-v-ii/2026-06-16weekly0.7 https://korshunov.ai/en/article/962-greed-is-learned-reward-channel-addiction-in-ai/2026-06-16weekly0.7 https://korshunov.ai/en/article/961-variance-in-llm-circuit-discovery-causes-and-mitigations/2026-06-16weekly0.7 https://korshunov.ai/en/article/960-ma-sbi-calibration-free-sbi-via-side-channel-guidance/2026-06-16weekly0.7 https://korshunov.ai/ru/article/961-razbros-v-otkrytii-skhem-llm-prichiny-i-mery-po-ustraneniiu/2026-06-16weekly0.7 https://korshunov.ai/en/article/959-raid-semantic-graph-diffusion-for-true-cold-start-and-cross-lingual-forecasting/2026-06-16weekly0.7 https://korshunov.ai/ru/article/959-raid-semanticheskoe-grafovoe-diffuzionnoe-modelirovanie-dlia-istinnogo-zapuska-i/2026-06-16weekly0.7 https://korshunov.ai/ru/article/960-ma-sbi-kalibrovka-bezopasnyi-sbi-cherez-napravlenie-storonnego-kanala/2026-06-16weekly0.7 https://korshunov.ai/en/article/958-unified-causal-origin-taxonomy-for-distributional-shifts-in-rl/2026-06-16weekly0.7 https://korshunov.ai/en/article/956-circuitlasso-scalable-circuit-learning-for-llm-interpretability/2026-06-16weekly0.7 https://korshunov.ai/ru/article/957-crossmaps-semanticheskaia-kartirovanie-s-uchetom-uverennosti-dlia-navigatsii/2026-06-16weekly0.7 https://korshunov.ai/ru/article/956-circuitlasso-masshtabiruemyi-metod-obucheniia-razrezhennykh-skhem-dlia-bolshikh/2026-06-16weekly0.7 https://korshunov.ai/en/article/957-crossmaps-confidence-aware-semantic-mapping-for-rover-navigation/2026-06-16weekly0.7 https://korshunov.ai/ru/article/958-edinaia-kausal-origin-topologiia-dlia-sdvigov-raspredeleniia-v-rl/2026-06-16weekly0.7 https://korshunov.ai/en/article/955-causal-model-of-theory-of-mind-in-ai-conflict/2026-06-16weekly0.7 https://korshunov.ai/en/article/953-low-frame-rate-degradation-in-neural-audio-codecs/2026-06-16weekly0.7 https://korshunov.ai/en/article/954-causal-framework-for-auditing-synthetic-data-disclosures/2026-06-16weekly0.7 https://korshunov.ai/ru/article/955-kauzalnaia-model-teorii-razuma-v-konflikte-ii/2026-06-16weekly0.7 https://korshunov.ai/ru/article/954-kauzalnyi-freimvork-dlia-audita-raskrytii-sinteticheskikh-dannykh/2026-06-16weekly0.7 https://korshunov.ai/ru/article/953-nizkaia-chastota-kadrov-v-neironnykh-audiokodekakh/2026-06-16weekly0.7 https://korshunov.ai/ru/article/951-dokumentatsiia-po-issledovaniiu-ii-uluchshaetsia-v-techenie-desiatiletiia/2026-06-16weekly0.7 https://korshunov.ai/en/article/951-ai-research-documentation-improves-over-decade/2026-06-16weekly0.7 https://korshunov.ai/ru/article/952-tekstovye-otzyvy-imeiut-ogranichennoe-vliianie-v-rekomendatelnykh-modeliakh/2026-06-16weekly0.7 https://korshunov.ai/ru/article/950-freimvork-agentnogo-llm-dlia-klassifikatsii-kodov-hts/2026-06-16weekly0.7 https://korshunov.ai/en/article/950-agentic-llm-framework-for-hts-code-classification/2026-06-16weekly0.7 https://korshunov.ai/en/article/952-textual-reviews-have-limited-impact-in-recommendation-models/2026-06-16weekly0.7 https://korshunov.ai/ru/article/949-progress-obuslovlennyi-ii-v-stabilnykh-meniu-obshchestvennykh-blag/2026-06-16weekly0.7 https://korshunov.ai/ru/article/948-pact-delegirovanie-malogo-iazykovogo-modeli-dlia-reaktivnogo-obucheniia-s/2026-06-16weekly0.7 https://korshunov.ai/en/article/948-pact-small-language-model-deliberation-for-reactive-reinforcement-learning/2026-06-16weekly0.7 https://korshunov.ai/en/article/949-ai-enabled-progress-in-stable-menus-of-public-goods/2026-06-16weekly0.7 https://korshunov.ai/en/article/946-bayesian-audits-reveal-inconsistent-ai-evaluation-timelines/2026-06-16weekly0.7 https://korshunov.ai/ru/article/945-tunejury-otkrytyi-metricheskii-instrument-dlia-vyravnivaniia-predpochtenii-v/2026-06-16weekly0.7 https://korshunov.ai/en/article/945-tunejury-open-metric-for-music-generation-preference-alignment/2026-06-16weekly0.7 https://korshunov.ai/en/article/947-activesam-fast-and-accurate-open-vocabulary-segmentation/2026-06-16weekly0.7 https://korshunov.ai/ru/article/947-activesam-bystryi-i-tochnyi-analiz-semanticheskoi-segmentatsii-s-otkrytym/2026-06-16weekly0.7 https://korshunov.ai/ru/article/946-baiesovskie-proverki-vyiavliaiut-nesoglasovannye-vremennye-linii-otsenki-ii/2026-06-16weekly0.7 https://korshunov.ai/ru/article/943-fusionrs-pervyi-masshtabnyi-dataset-rgb-infrakrasnogo-distantsionnogo/2026-06-16weekly0.7 https://korshunov.ai/ru/article/944-tokenpilot-effektivnoe-upravlenie-kontekstom-dlia-agentov-llm/2026-06-16weekly0.7 https://korshunov.ai/ru/article/942-hamon-passivnaia-opticheskaia-predskazatelnaia-sistema-dlia-prognozirovaniia-na/2026-06-16weekly0.7 https://korshunov.ai/en/article/942-hamon-passive-optical-forecasting-for-long-horizon-time-series/2026-06-16weekly0.7 https://korshunov.ai/en/article/944-tokenpilot-cache-efficient-context-management-for-llm-agents/2026-06-16weekly0.7 https://korshunov.ai/en/article/943-fusionrs-first-large-scale-rgb-infrared-remote-sensing-dataset/2026-06-16weekly0.7 https://korshunov.ai/ru/article/940-satia-nadella-o-loopcraft-i-ekosistemakh-frontier/2026-06-16weekly0.7 https://korshunov.ai/en/article/1005-hyperball-optimization-for-faster-language-model-training/2026-06-16weekly0.7 https://korshunov.ai/en/article/941-phase-in-neural-representations-an-internal-oppenheim-lim-test/2026-06-16weekly0.7 https://korshunov.ai/ru/article/941-faza-v-neironnykh-predstavleniiakh-vnutrennii-test-oppengeima-lima/2026-06-16weekly0.7 https://korshunov.ai/en/article/940-satya-nadella-on-loopcraft-and-frontier-ecosystems/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1005-optimizatsiia-hyperball-dlia-uskoreniia-obucheniia-iazykovykh-modelei/2026-06-16weekly0.7 https://korshunov.ai/en/article/1004-factorized-neural-operators-decompose-dynamic-and-persistent-responses/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1003-ceap-snizhaet-dispersiiu-v-otkrytii-tsirkuitov-v-llm/2026-06-16weekly0.7 https://korshunov.ai/en/article/1002-adaptive-functional-gradient-descent-with-convergence-guarantees/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1002-adaptivnyi-funktsionalnyi-gradientnyi-spusk-s-garantiiami-skhodimosti/2026-06-16weekly0.7 https://korshunov.ai/en/article/1003-ceap-reduces-variance-in-llm-circuit-discovery/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1004-faktorizovannye-neironnye-operatory-razbivaiut-dinamicheskie-i-ustoichivye/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1001-edinaia-kausalnaia-klassifikatsiia-istochnikov-smeshchenii-raspredelenii-v-rl/2026-06-16weekly0.7 https://korshunov.ai/en/article/1000-key-properties-for-effective-code-interpreter-reasoning/2026-06-16weekly0.7 https://korshunov.ai/ru/article/999-crossmaps-semanticheskaia-kartirovanie-s-uchetom-uverennosti-dlia-navigatsii/2026-06-16weekly0.7 https://korshunov.ai/en/article/999-crossmaps-confidence-aware-semantic-mapping-for-rover-navigation/2026-06-16weekly0.7 https://korshunov.ai/ru/article/1000-kliuchevye-svoistva-dlia-effektivnogo-rassuzhdeniia-kod-interpretatora/2026-06-16weekly0.7 https://korshunov.ai/en/article/1001-unified-causal-origin-taxonomy-of-distributional-shifts-in-rl/2026-06-16weekly0.7 https://korshunov.ai/en/article/997-a-nonparametric-two-sample-test-using-prelu-ipm/2026-06-16weekly0.7 https://korshunov.ai/ru/article/995-kauzalnyi-podkhod-k-auditu-raskrytii-sinteticheskikh-dannykh/2026-06-16weekly0.7 https://korshunov.ai/ru/article/997-neparametricheskii-dvukhvyborochnyi-test-s-ispolzovaniem-prelu-ipm/2026-06-16weekly0.7 https://korshunov.ai/en/article/998-circuitlasso-scalable-circuit-learning-for-llm-interpretability/2026-06-16weekly0.7 https://korshunov.ai/en/article/995-causal-framework-for-auditing-synthetic-data-disclosures/2026-06-16weekly0.7 https://korshunov.ai/ru/article/998-circuitlasso-masshtabiruemoe-obuchenie-skhem-dlia-interpretiruemosti-llm/2026-06-16weekly0.7 https://korshunov.ai/en/article/996-latent-space-mapping-for-interpretable-molecular-coordinates-from-nanopore/2026-06-16weekly0.7 https://korshunov.ai/ru/article/994-gibridnaia-konvoliutsionnaia-vae-dlia-kriptovaliutnykh-poverkhnostei/2026-06-16weekly0.7 https://korshunov.ai/en/article/994-hybrid-convolutional-vae-for-crypto-volatility-surfaces/2026-06-16weekly0.7 https://korshunov.ai/ru/article/996-mapping-latentnogo-prostranstva-dlia-interpretiruemoi-molekuliarnoi-koordinaty/2026-06-16weekly0.7 https://korshunov.ai/en/article/993-fixed-size-neural-networks-achieve-arbitrary-sobolev-approximation/2026-06-16weekly0.7 https://korshunov.ai/en/article/991-spaik-scalable-pairwise-kernel-learning-with-stochastic-vec-trick/2026-06-16weekly0.7 https://korshunov.ai/ru/article/993-neironnye-seti-fiksirovannogo-razmera-dostigaiut-proizvolnogo-priblizheniia-v/2026-06-16weekly0.7 https://korshunov.ai/ru/article/992-metod-rezidualnogo-obucheniia-s-ispolzovaniem-napravliaiushchego-kontrolia-dlia/2026-06-16weekly0.7 https://korshunov.ai/ru/article/991-spaik-masshtabiruemoe-obuchenie-parnykh-iader-s-pomoshchiu-statisticheskogo/2026-06-16weekly0.7 https://korshunov.ai/en/article/992-task-error-residual-learning-for-real-robot-five-ball-juggling/2026-06-16weekly0.7 https://korshunov.ai/en/article/988-fingerprinting-agent-behavior-through-procedural-trajectories/2026-06-16weekly0.7 https://korshunov.ai/ru/article/990-statisticheskaia-uproshchenie-razdeliaet-inferens-ot-obnovleniia-sostoianiia/2026-06-16weekly0.7 https://korshunov.ai/en/article/990-probabilistic-thinning-decouples-inference-from-state-updates/2026-06-16weekly0.7 https://korshunov.ai/ru/article/988-opredelenie-povedeniia-agentov-cherez-protsedury-traektorii/2026-06-16weekly0.7 https://korshunov.ai/ru/article/989-dynestyx-veroiatnostnoe-programmirovanie-dlia-dinamicheskikh-sistem/2026-06-16weekly0.7 https://korshunov.ai/en/article/989-dynestyx-probabilistic-programming-for-dynamical-systems/2026-06-16weekly0.7 https://korshunov.ai/ru/article/985-pact-delegirovanie-malogo-iazykovogo-modeli-dlia-reaktivnogo-obucheniia-s/2026-06-16weekly0.7 https://korshunov.ai/en/article/985-pact-small-language-model-deliberation-for-reactive-reinforcement-learning/2026-06-16weekly0.7 https://korshunov.ai/ru/article/987-analiticheskaia-torsiia-i-pogloshchenie-spektralnogo-razryva-v-proizvoditelnosti/2026-06-16weekly0.7 https://korshunov.ai/ru/article/986-mnogotsentrovyi-benchmark-dlia-diagnostiki-zabolevanii-briushnoi-polosti-na-ne/2026-06-16weekly0.7 https://korshunov.ai/en/article/986-multi-center-benchmark-for-abdominal-disease-diagnosis-from-non-contrast-ct/2026-06-16weekly0.7 https://korshunov.ai/en/article/987-analytic-torsion-and-spectral-gap-capture-persistent-laplacian-performance/2026-06-16weekly0.7 https://korshunov.ai/ru/article/984-activesam-bystryi-i-tochnyi-otkrytie-slovarnyi-segmentatsiia/2026-06-16weekly0.7 https://korshunov.ai/ru/article/983-posle-posledovatelnye-operatory-lozhnoi-falsifikatsii-ne-uluchshaiut-tochnost-v/2026-06-16weekly0.7 https://korshunov.ai/en/article/984-activesam-fast-and-accurate-open-vocabulary-segmentation/2026-06-16weekly0.7 https://korshunov.ai/ru/article/982-ppad-slozhnost-dlia-min-maks-optimizatsii-kvadratichnykh-polinomov/2026-06-16weekly0.7 https://korshunov.ai/en/article/982-ppad-hardness-for-min-max-optimization-of-quadratic-polynomials/2026-06-16weekly0.7 https://korshunov.ai/en/article/983-post-hoc-falsification-operators-fail-to-improve-accuracy-in-small-code-models/2026-06-16weekly0.7 https://korshunov.ai/ru/article/979-rove-obuchenie-s-pomoshchiu-vmeshatelstv-cheloveka-dlia-manipuliatsii-robota/2026-06-16weekly0.7 https://korshunov.ai/en/article/980-neural-exposure-interaction-search-for-interpretable-hte/2026-06-16weekly0.7 https://korshunov.ai/ru/article/980-neironnaia-model-vzaimodeistviia-ekspozitsii-dlia-interpretiruemogo-effekta/2026-06-16weekly0.7 https://korshunov.ai/en/article/981-tunejury-open-metric-for-music-generation-preference-alignment/2026-06-16weekly0.7 https://korshunov.ai/en/article/979-rove-reinforcement-learning-with-human-interventions-for-humanoid-manipulation/2026-06-16weekly0.7 https://korshunov.ai/ru/article/981-tunejury-otkrytyi-metricheskii-instrument-dlia-vyravnivaniia-predpochtenii-v/2026-06-16weekly0.7 https://korshunov.ai/en/article/977-filtered-conformal-ellipsoids-for-graph-native-time-series/2026-06-16weekly0.7 https://korshunov.ai/en/article/976-tokenpilot-cache-efficient-context-management-for-llm-agents/2026-06-16weekly0.7 https://korshunov.ai/ru/article/978-ostatochnye-soedineniia-smiagchaiut-problemy-gradientov-v-glubokikh-setiakh/2026-06-16weekly0.7 https://korshunov.ai/ru/article/977-filtrovannye-konformnye-ellipsoidy-dlia-graf-orientirovannykh-vremennykh-riadov/2026-06-16weekly0.7 https://korshunov.ai/ru/article/976-tokenpilot-effektivnoe-upravlenie-kontekstom-dlia-agentov-llm/2026-06-16weekly0.7 https://korshunov.ai/en/article/978-residual-connections-mitigate-gradient-issues-in-deep-networks/2026-06-16weekly0.7 https://korshunov.ai/ru/article/975-matematicheskii-obzor-analiza-prostranstva-form-v-mashinnom-obuchenii/2026-06-16weekly0.7 https://korshunov.ai/ru/article/973-hamon-passivnaia-opticheskaia-sistema-prognozirovaniia/2026-06-16weekly0.7 https://korshunov.ai/en/article/973-hamon-passive-optical-forecasting-core/2026-06-16weekly0.7 https://korshunov.ai/en/article/975-a-mathematical-review-of-shape-space-analysis-in-machine-learning/2026-06-16weekly0.7 https://korshunov.ai/ru/article/974-exprl-issledovatelskaia-rl-dlia-srednego-obucheniia-llm/2026-06-16weekly0.7 https://korshunov.ai/en/article/974-exprl-exploratory-rl-for-llm-mid-training/2026-06-16weekly0.7 https://korshunov.ai/ru/article/970-faza-v-neironnykh-predstavleniiakh-vnutrennii-test-oppengeima-lima/2026-06-16weekly0.7 https://korshunov.ai/ru/article/971-ataka-ring-ispolzovanie-differentsialnoi-konfidentsialnosti-v-raspredelennom/2026-06-16weekly0.7 https://korshunov.ai/en/article/972-kveraser-efficient-localized-context-erasing-in-llms/2026-06-16weekly0.7 https://korshunov.ai/en/article/970-phase-in-neural-representations-an-internal-oppenheim-lim-test/2026-06-16weekly0.7 https://korshunov.ai/ru/article/972-kveraser-effektivnoe-lokalnoe-udalenie-konteksta-v-llms/2026-06-16weekly0.7 https://korshunov.ai/en/article/971-dp-fl-backdoor-attacks-ring-exploits-privacy-for-malicious-signals/2026-06-16weekly0.7 https://korshunov.ai/ru/article/969-habc-uluchshaet-rl-nastroiku-vlas-s-razrezhennymi-rezultatami/2026-06-16weekly0.7 https://korshunov.ai/en/article/967-exact-posterior-score-estimation-for-linear-inverse-problems/2026-06-16weekly0.7 https://korshunov.ai/ru/article/967-tochnaia-otsenka-posteriornogo-skoora-dlia-lineinykh-obratnykh-zadach/2026-06-16weekly0.7 https://korshunov.ai/ru/article/968-geometricheskaia-model-deistvii-dlia-obucheniia-politik-robotov/2026-06-16weekly0.7 https://korshunov.ai/en/article/968-geometric-action-model-for-robot-policy-learning/2026-06-16weekly0.7 https://korshunov.ai/en/article/969-habc-improves-rl-fine-tuning-of-vlas-with-sparse-outcomes/2026-06-16weekly0.7 https://korshunov.ai/en/article/937-qwable-v1-released-as-distillation-of-claude-fable-5/2026-06-16weekly0.7 https://korshunov.ai/ru/article/937-vypushchen-qwable-v1-kak-distilliat-claude-fable-5/2026-06-16weekly0.7 https://korshunov.ai/ru/article/934-vllm-vypustil-novyi-parser-potoka-dlia-qwen3-v-nochnoi-versii/2026-06-16weekly0.7 https://korshunov.ai/en/article/934-vllm-releases-new-streaming-parser-for-qwen3-in-nightly/2026-06-16weekly0.7 https://korshunov.ai/en/article/931-cloudflare-captcha-triggered-only-for-searches-with-ampersand/2026-06-16weekly0.7 https://korshunov.ai/ru/article/933-halbench-proveriaet-29-otkrytykh-modelei-na-sikofantizm-i-khalliutsinatsii/2026-06-16weekly0.7 https://korshunov.ai/ru/article/931-cloudflare-captcha-aktiviruetsia-tolko-dlia-zaprosov-s-ampersandom/2026-06-16weekly0.7 https://korshunov.ai/en/article/933-halbench-tests-29-open-source-models-on-sycophancy-and-hallucination/2026-06-16weekly0.7 https://korshunov.ai/en/article/930-gemma3-270m-model-released-on-reddit/2026-06-16weekly0.7 https://korshunov.ai/en/article/929-datasette-agent-0-3a0-releases-with-user-approval-for-write-sql-operations/2026-06-16weekly0.7 https://korshunov.ai/ru/article/929-datasette-agent-0-3a0-vypuskaetsia-s-soglasiia-polzovatelia-po-operatsiiam-sql/2026-06-16weekly0.7 https://korshunov.ai/ru/article/930-vypushchen-model-gemma3-270m-na-reddit/2026-06-16weekly0.7