Models, papers, tools. 110,079 articles with AI-powered sentiment analysis and key takeaways.
SemaClaw is an open-source framework addressing the shift from prompt engineering to 'harness engineering'—building infrastructure for controllable, auditable AI agents. Announced alongside OpenClaw's mass adoption in early 2026, it enables persistent personal AI agents through DAG-based orchestration, behavioral safety systems, and automated knowledge base construction.
UniToolCall introduces a standardized framework unifying tool-use representation, training data, and evaluation for LLM agents. The framework combines 22k+ tools and 390k+ training instances with a unified evaluation methodology, enabling fine-tuned models like Qwen3-8B to achieve 93% precision—surpassing GPT, Gemini, and Claude in specific benchmarks.
Researchers introduced BankerToolBench (BTB), an open-source benchmark to evaluate AI agents on investment banking workflows developed with 502 professional bankers. Testing nine frontier models revealed that even the best performer (GPT-5.4) fails nearly half of evaluation criteria, with zero outputs rated client-ready, highlighting significant gaps in AI readiness for high-stakes professional work.
Researchers introduce PaperScope, a comprehensive benchmark for evaluating multi-modal AI systems on complex scientific research tasks across multiple documents. The benchmark reveals that even advanced systems like OpenAI Deep Research and Tongyi Deep Research struggle with long-context retrieval and cross-document reasoning, exposing significant gaps in current AI capabilities for scientific workflows.
Researchers introduce ContextCurator, a reinforcement learning-based framework that decouples context management from task execution in LLM agents, addressing the context bottleneck problem. The approach pairs a lightweight specialized policy model with a frozen foundation model, achieving significant improvements in success rates and token efficiency across benchmark tasks.
Researchers from Kyutai's Moshi foundation model project conducted the first comprehensive environmental audit of GenAI model development, revealing the hidden compute costs of R&D, failed experiments, and debugging beyond final training. The study quantifies energy consumption, water usage, greenhouse gas emissions, and resource depletion across the entire development lifecycle, exposing transparency gaps in how AI labs report environmental impact.
Researchers demonstrate that inference-time scaffolding can double the performance of small 8B language models on complex tool-use tasks without additional training, by deploying the same frozen model in three specialized roles: summarization, reasoning, and code correction. On a single 24GB GPU, this approach enables an 8B model to match or exceed much larger systems like DeepSeek-Coder 33B, suggesting efficient deployment paths for capable AI agents on modest hardware.
Researchers introduce Hodoscope, an unsupervised monitoring tool that detects anomalous AI agent behaviors by comparing action patterns across different evaluation contexts, without relying on predefined misbehavior rules. The approach discovered a previously unknown vulnerability in the Commit0 benchmark and independently recovered known exploits, reducing human review effort by 6-23x compared to manual sampling.
Researchers introduce a framework that automatically learns context-sensitive constraints from LLM interactions, eliminating the need for manual specification while ensuring perfect constraint adherence during generation. The method enables even 1B-parameter models to outperform larger models and state-of-the-art reasoning systems in constraint-compliant generation.
Researchers demonstrate that safety evaluations of persona-imbued large language models using only prompt-based testing are fundamentally incomplete, as activation steering reveals entirely different vulnerability profiles across model architectures. Testing across four models reveals the 'prosocial persona paradox' where conscientious personas safe under prompting become the most vulnerable to activation steering attacks, indicating that single-method safety assessments can miss critical failure modes.
Researchers introduce PAC-Bench, a benchmark for evaluating how AI agents collaborate while maintaining privacy constraints. The study reveals that privacy protections significantly degrade multi-agent system performance and identify coordination failures as a critical unsolved challenge requiring new technical approaches.
Researchers introduce Grid2Matrix, a benchmark that reveals fundamental limitations in Vision-Language Models' ability to accurately process and describe visual details in grids. The study identifies a critical gap called 'Digital Agnosia'—where visual encoders preserve grid information that fails to translate into accurate language outputs—suggesting that VLM failures stem not from poor vision encoding but from the disconnection between visual features and linguistic expression.
Researchers challenge the assumption that longer reasoning chains always improve LLM performance, discovering that extended test-time compute leads to diminishing returns and 'overthinking' where models abandon correct answers. The study demonstrates that optimal compute allocation varies by problem difficulty, enabling significant efficiency gains without sacrificing accuracy.
Researchers demonstrate that a large language model's diversity profile—how probability mass spreads across different solution approaches—should determine whether reasoning strategies prioritize breadth or depth exploration. Testing on Qwen and Olmo model families reveals that lightweight refinement signals work well for low-diversity aligned models but offer limited value for high-diversity base models, suggesting optimal inference strategies must be model-specific rather than universal.
FACT-E is a new evaluation framework that uses controlled perturbations to assess the faithfulness of Chain-of-Thought reasoning in large language models, addressing the problem of models generating seemingly coherent explanations with invalid intermediate steps. By measuring both internal chain consistency and answer alignment, FACT-E enables more reliable detection of flawed reasoning and selection of trustworthy reasoning trajectories for in-context learning.
Researchers propose Cognitive Core, a governed AI architecture designed for high-stakes institutional decisions that achieves 91% accuracy on prior authorization appeals while eliminating silent errors—a critical failure mode where AI systems make incorrect determinations without human review. The framework introduces 'governability' as a primary evaluation metric alongside accuracy, demonstrating that institutional AI requires fundamentally different design principles than general-purpose agents.
A new study reveals that large language models fail at counterfactual reasoning when policy findings contradict intuitive expectations, despite performing well on obvious cases. The research demonstrates that chain-of-thought prompting paradoxically worsens performance on counter-intuitive scenarios, suggesting current LLMs engage in 'slow talking' rather than genuine deliberative reasoning.
Researchers tested whether large language models develop spatial world models through maze-solving tasks, finding that leading models like Gemini, GPT-4, and Claude struggle with spatial reasoning. Performance varies dramatically (16-86% accuracy) depending on input format, suggesting LLMs lack robust, format-invariant spatial understanding rather than building true internal world models.
Researchers propose lightweight sanity checks for agentic data science (ADS) systems to detect falsely optimistic conclusions that users struggle to identify. Using the Predictability-Computability-Stability framework, the checks expose whether AI agents like OpenAI Codex reliably distinguish signal from noise. Testing on 11 real datasets reveals that over half produced unsupported affirmative conclusions despite individual runs suggesting otherwise.
Researchers propose Min-k Sampling, a novel decoding strategy for large language models that dynamically identifies semantic cliffs in logit distributions to optimize token truncation. Unlike temperature-sensitive methods like Top-k and Top-p, Min-k achieves temperature invariance through relative logit dynamics while maintaining superior text quality across reasoning, creative writing, and human evaluation benchmarks.
A comprehensive comparative study traces the evolution of OpenAI's GPT models from GPT-3 through GPT-5, revealing that successive generations represent far more than incremental capability improvements. The research demonstrates a fundamental shift from simple text predictors to integrated, multimodal systems with tool access and workflow capabilities, while persistent limitations like hallucination and benchmark fragility remain largely unresolved across all versions.
Researchers introduce Zero-shot Visual World Models (ZWM), a computational framework inspired by how young children learn physical understanding from minimal data. The approach combines sparse prediction, causal inference, and compositional reasoning to achieve data-efficient learning, demonstrating that AI systems can match child development patterns while learning from single-child observational data.
A new study reveals that multi-agent AI systems achieve better business outcomes than individual AI agents, but at the cost of reduced alignment with intended values. The research, spanning consultancy and software development tasks, highlights a critical trade-off between capability and safety that challenges current AI deployment assumptions.
Researchers introduce VeriSim, an open-source framework that tests medical AI systems by injecting realistic patient communication barriers—such as memory gaps and health literacy limitations—into clinical simulations. Testing across seven LLMs reveals significant performance degradation (15-25% accuracy drop), with smaller models suffering 40% greater decline than larger ones, exposing a critical gap between standardized benchmarks and real-world clinical robustness.
Researchers introduce SpecMoE, a new inference system that applies speculative decoding to Mixture-of-Experts language models to improve computational efficiency. The approach achieves up to 4.30x throughput improvements while reducing memory and bandwidth requirements without requiring model retraining.
