Xcientist introduces a research harness that externalizes and makes visible the reasoning processes in AI scientists. It preserves traceable, contract-governed research artifacts from problem formulation to validation, addressing claim drift and ensuring scientific accountability.
BeliefDiffusion combines diffusion models for multimodal belief representation with Model Predictive Control for long-term navigation planning. It outperforms model-free reinforcement learning and other generative methods in navigation success and path efficiency in partially observable environments.
SGCD introduces an iterative framework to improve GUI agents by addressing supervision gaps in off-trajectory states. It extracts skills from both successful and failed rollouts, using them to guide policy continuations that are mixed with expert trajectories. On OSWorld-Verified, SGCD boosts success rates of three base models from low-30\% to over 50\%.
Decoupled Search Grounding (DSG) separates search functionality from reasoning models, enabling vendor-agnostic, tunable, and reusable search grounding. DSG achieves near-native accuracy on SimpleQA with 91% lower search cost and 99.4% warm-cache hit rate, while reducing latency by 68% and preserving concise output contracts.
RTSGameBench addresses limitations in existing RTS benchmarks by offering diverse gameplay, targeted competency diagnosis, and self-evolving scenario generation. It evaluates vision-language models in strategic reasoning under uncertainty, revealing that state-of-the-art models struggle with multiagent coordination and large-scale tasks.
CAPRA is a multi-agent LLM system that generates personalized, template-compliant LaTeX feedback on software architecture deliverables. It uses specialized agents, PyMuPDF, and gpt-4o to extract and analyze text and UML diagrams, with evidence anchoring and consistency management to ensure reliability. A preliminary evaluation of 10 student reports shows CAPRA met 88.8% of eight criteria and achieved moderate inter-rater agreement (kappa = 0.582), with each report processed in under 4 minutes.
ThinkDeception introduces a progressive reinforcement learning framework that enables interpretable multimodal deception detection. It leverages a step-by-step annotated Chain of Thought dataset and proposes Visual-Audio Consistency Group Relative Policy Optimization with a dynamic curriculum, enhancing reasoning quality and outperforming existing methods on mainstream benchmarks.
TRAP evaluates how well models complete tasks using private data without leaking it. Across 22 models, all show non-trivial privacy leakage, with instruction-following ability linked to higher leakage. Structural private field isolation prevents leakage by replacing private fields with hash keys, maintaining task accuracy without sacrificing privacy.
Spotlight enables DiT RL post-training by leveraging idle spot GPUs, reducing costs by 1.4-6.4× while achieving superior image quality. It uses stale model weights in exploration and reconfigures sequence parallelism in real time, allowing efficient GPU utilization without breaking training pipelines.
RODS addresses sample depletion in multi-turn tool-use RL by using reward variance to detect capability boundaries. It synthesizes new data in real time, matching structural complexity of boundary samples, and maintains a dynamic replay buffer that co-evolves with the policy. RODS achieves performance comparable to a 17K-sample offline pipeline with 20x fewer trajectories.
ARIADNE enables dynamic, training-free adapter selection at inference time by using centroids from adapter training data embeddings. It selects the most appropriate adapter based on proximity in latent space, without requiring access to adapter internals or additional training, and achieves 89.7% average selection accuracy across 44 NLP tasks.
A study finds that leadership styles in multi-agent LLM teams only improve performance when the initial consensus is unreliable, recoverable, and not self-corrected by undirected interaction. Process-level coordination control adds value only under specific conditions predicted by team science, with no single leadership style outperforming others in accuracy across tasks and models.
A new paper proposes a fundamental redesign of the web to prioritize AI agent access, challenging the long-held assumption that humans are the primary web users. It introduces access, economic, and content layer reforms—including agent-identifiable HTTP headers, intent-based subscription models, and a cryptographic provenance system—to enable AI agents as first-class participants, with human supervision and accountability embedded in the architecture.
A new taxonomy classifies LLM agent communication protocols across five dimensions: counterparty, payload, interaction state, discovery mechanism, and schema flexibility. Analysis shows hybrid payloads, session-state persistence, and runtime schema negotiation are common, with decentralized discovery remaining rare. The study predicts short-term convergence toward unified agent-to-agent and agent-to-context protocols, and long-term evolution toward a federated, layered protocol stack.
The Human-AI Coevolution Dynamics Framework (HACD-H) introduces a unified model for long-term human-AI interaction, integrating emotional adaptation, memory, and personality into a self-organizing system. Results show social intelligence emerges through coevolution, with a significant negative correlation between social intelligence and social cognitive energy (r = -0.391, p < 0.001), and progressive energy reduction over time.
AdsMind is a closed-loop multi-agent system that uses machine learning force fields and feedback to correct errors in adsorption configuration searches on catalyst surfaces. It achieves 100% and 98.8% success rates on AA20 and OCD-GMAE62 benchmarks, reduces energy dispersion by 14-fold compared to baselines, and maintains correct adsorption-energy signs in DFT validation, outperforming open-loop LLM agents.
A user with a 12GB VRAM GPU asks for model recommendations for general chatting, roleplaying, and coding. They prioritize uncensored models for chat and roleplaying, and have a Ryzen 5600 CPU and 32GB DDR4 RAM.
Lemonade v10.8 introduces dynamic VRAM management that auto-unloads idle models and downsizes KV-cache to reclaim GPU memory. It adds cloud offload support for OpenAI-compatible providers, enabling local-first model serving with optional cloud routing. A new MCP gateway exposes local models as tools via POST /mcp, allowing local models to be used as tools in MCP-aware applications.
A user trained a language model to roll a die, ensuring each number appears approximately once every six rolls. The post highlights how mainstream LLMs tend to default to saying '4' when asked to roll a die, illustrating a broader issue in reinforcement learning: models often fail to explore effectively and instead follow known patterns.
Radical AI has accelerated materials discovery by producing and characterizing 1,200 alloys in six months—nearly 10x faster than DARPA/GE MACH's goal of 500 alloys in a year. Their self-driving labs use AI scientists to generate and test hypotheses in closed-loop systems, leading to 300 new materials with 10 exhibiting novel, state-of-the-art properties now being developed for commercial use.