The llama.cpp project has released version b9844, which introduces ggml-webgpu support for the NVFP4 quantization format. This update also provides pre-built binaries for macOS, iOS, Linux, Android, Windows, and openEuler across various hardware backends.
This study evaluates the ability of large-language models to approximate human cultural tastes by generating silicon surrogates from the Survey of Public Participation in the Arts. Using models from OpenAI, Anthropic, and DeepSeek, the authors analyze 277,470 synthetic respondents to determine if LLMs can faithfully replicate real-world survey data.
Researchers propose TIGRAG (Token-Induced GraphRAG), a framework that uses token co-occurrence statistics to build scalable knowledge graphs for efficient retrieval-augmented generation. This approach addresses the limitations of standard RAG in multi-hop reasoning by avoiding expensive LLM-based extraction pipelines.
Researchers introduce an information-theoretic framework to quantify the directed flow of semantic content between interlocutors and decompose multi-source contributions into redundant, unique, and synergistic components.
This study investigates whether verbose chain-of-thought prompting improves large language model reasoning through increased computation or by providing useful semantic content. The authors present evidence from in-distribution sampling and controlled interventions to determine the specific factors driving performance gains.
This study evaluates the performance gains of transformer-based DNA language models like DNABERT2 compared to conventional approaches such as ConvNova, specifically addressing the high cost of pre-training. It investigates whether these improvements justify the computational overhead and analyzes the impact of Byte Pair Encoding (BPE) tokenization on genomic tasks.
This study addresses the conflation of grammatical gender and social semantic bias in contextual language models for gendered languages like Spanish, proposing a framework to disentangle these dimensions. The authors construct balanced datasets using controlled templates and natural Wikipedia contexts to estimate gender directions while suppressing contamination.
The authors introduce Cortex, a framework that transforms web-scale corpus construction from flat document filtering into structured knowledge organization using an Ontological Corpus Graph (OCG). This three-layer structure unifies quality-refined content, hierarchical lightweight ontology, and cross-domain alignment to address the escalating data requirements of large language models.
Researchers introduce the Dynamic Agent-based Interaction Network (DAIN), a framework that reconceptualizes multimodal fusion as a dynamic, multi-agent collaborative process rather than relying on static architectures. DAIN utilizes a context-aware Meta-Controller to dynamically schedule sparse activation of specialized agents and orchestrates compressed communication for consensus-building.
This paper analyzes non-sequential multimodal sentence-level embeddings, focusing on the SONAR model, to demonstrate that specific embedding dimensions are sensitive to perturbations and can indicate decoding anomalies. By leveraging consistency between successive encoding and decoding, the authors successfully build an accurate anomaly detector.
The authors propose PRP, a Proactive Routing Paradigm that accelerates inference in large multimodal models by enabling early decision-making through joint evaluation of draft and target model competence. This approach addresses the bottleneck of establishing reliable query difficulty signals in multimodal settings without relying on data-sensitive supervised fine-tuning or post-hoc token probabilities.
This paper addresses the shared measurement problem in LLM evaluation and AI safety, where benchmark scores often improve while latent safety properties remain difficult to verify. It introduces EvalSafetyGap, a hybrid survey and conceptual framework combining systematic evidence synthesis with a structured audit of ten models.
This study evaluates the use of domain-specific transformer embeddings combined with classical machine learning models to detect dosing errors in clinical trial protocols. The research aims to improve patient safety and trial integrity by identifying preventable medication errors early through text representation analysis.
This study compared the recognition performance of human listeners against three state-of-the-art off-the-shelf ASR systems (Whisper-large-V3, Google Chirp 3, and Omnilingual) on Dutch continuous read and spontaneous speech from a single speaker with severe dysarthria.
This article presents a theoretical framework for grounding large language model reasoning trajectories when relying on incomplete knowledge graph evidence rather than complete truth states.
This article proposes a novel multi-agent system that emulates human annotator decision-making processes to detect and debunk disinformation, achieving superior results compared to individual Large Language Models like GPT-4 and GPT-3.5.
This article develops a theory for speculative decoding regimes that use greedy decoding, relaxed acceptance rules, or tree-based candidate sets, rather than the stochastic distribution-preserving settings studied in existing literature. The authors characterize rejection regions as lower level sets of the target distribution to derive exact KL divergence requirements and sharp margin-based bounds for various acceptance criteria.
Researchers present DialogPII, a multilingual dataset of synthetic dialog transcripts designed to support the development and evaluation of automatic systems for detecting personally identifiable information. This resource addresses privacy concerns in sensitive domains by providing annotated data across 11 languages and eight interaction scenarios.
The authors propose a novel training approach for end-to-end automatic speech recognition (ASR) that addresses noisy labels and lack of domain specificity in large-scale weakly supervised datasets. The method involves pretraining on the full dataset, continued pretraining on a filtered subset based on character error rate, and fine-tuning on acoustically similar samples from that subset.
A user tested Qwen3.6-27B (8-bit) alongside GLM5.2 using a coding harness that employs three critics—code review, test review, and Playwright e2e—to validate output quality.