06:00AI models struggle to explain chess and Go moves in natural languageResearch & Sciencecs.AI updates on arXiv.org8/10Relevance 8 out of 10
06:00HyperPersona: A hypergraph framework for text-based personality predictionResearch & Sciencecs.AI updates on arXiv.org8/10Relevance 8 out of 10
06:00Strategies for simulator recalibration under data driftResearch & Sciencecs.AI updates on arXiv.org7/10Relevance 7 out of 10
06:00Tree-Aware Posterior Distillation for Long-Horizon Agentic Reinforcement LearningResearch & Sciencecs.AI updates on arXiv.org8/10Relevance 8 out of 10
06:00MoEvolve: A distillation method to strengthen MoE modelsResearch & Sciencecs.AI updates on arXiv.org9/10Relevance 9 out of 10
06:00LangNCO: Compiling Natural-Language Constraints into Inference-Time Interventions for Neural Combinatorial OptimizationResearch & Sciencecs.AI updates on arXiv.org9/10Relevance 9 out of 10
06:00Hierarchical model of perceptual learning for spatial hearing based on free energy principleResearch & Sciencecs.AI updates on arXiv.org7/10Relevance 7 out of 10
06:00Moral sycophancy in vision-language modelsResearch & Sciencecs.AI updates on arXiv.org8/10Relevance 8 out of 10
06:00MGE-PO: Entropy-Aware Credit Assignment for Mathematical ReasoningResearch & Sciencecs.AI updates on arXiv.org8/10Relevance 8 out of 10
06:00FlashReFocus: Interactive Image Refocusing in SecondsResearch & Sciencecs.AI updates on arXiv.org9/10Relevance 9 out of 10
06:00Evaluating LLM-based enterprise data assistants with knowledge graphsResearch & Sciencecs.AI updates on arXiv.org8/10Relevance 8 out of 10
06:00MobileInject: New benchmark for testing MLLM-based mobile agents against multimodal injection attacksResearch & Sciencecs.AI updates on arXiv.org9/10Relevance 9 out of 10
06:00Language models learn to improve other language modelsResearch & Sciencecs.AI updates on arXiv.org9/10Relevance 9 out of 10
06:00AWAM: Active Trajectory-Conditioned View Imagination for Robust Robot ManipulationResearch & Sciencecs.AI updates on arXiv.org9/10Relevance 9 out of 10
06:00RECREG: A framework for regular safety properties in reinforcement learningResearch & Sciencecs.AI updates on arXiv.org8/10Relevance 8 out of 10
06:00QD-Rec: Open-Ended Recommendations via Quality-DiversityResearch & Sciencecs.AI updates on arXiv.org8/10Relevance 8 out of 10
06:00A topological signature of learning with noisy labelsResearch & Sciencecs.AI updates on arXiv.org8/10Relevance 8 out of 10
06:00Factual Confidence Estimation in LLMs via Contrastive LikelihoodsResearch & Sciencecs.AI updates on arXiv.org8/10Relevance 8 out of 10
06:00Evaluation of small language models for malicious code generationResearch & Sciencecs.AI updates on arXiv.org8/10Relevance 8 out of 10
06:00Reason-guided human reward modeling for preference-based reinforcement learningResearch & Sciencecs.AI updates on arXiv.org8/10Relevance 8 out of 10
06:00SIT-Eval: Interactive evaluation of conversational agents through situational scenariosResearch & Sciencecs.AI updates on arXiv.org9/10Relevance 9 out of 10
06:00Search-Free Planning in Latent Space for Visual ControlResearch & Sciencecs.AI updates on arXiv.org8/10Relevance 8 out of 10
06:00PresenceBench: Evaluating LLM Adaptation to User StateResearch & Sciencecs.AI updates on arXiv.org9/10Relevance 9 out of 10
06:00TUPLE-Bench: Evaluating privacy-utility trade-offs in typed decision modelsResearch & Sciencecs.AI updates on arXiv.org8/10Relevance 8 out of 10
06:00Neural Expansion: A Unified Mechanism for Deep Neural Network GeneralizationResearch & Sciencecs.AI updates on arXiv.org9/10Relevance 9 out of 10
06:00Behavioral Couplings: Unintended Behaviors Emerge in Fine-Tuned LLMsResearch & Sciencecs.AI updates on arXiv.org9/10Relevance 9 out of 10
06:00New adaptive method for pixel diffusion transformers with MFlowResearch & Sciencecs.AI updates on arXiv.org8/10Relevance 8 out of 10
06:00GVPO: A new optimization method for LLM post-trainingResearch & Sciencecs.AI updates on arXiv.org9/10Relevance 9 out of 10
06:00Survey of Explainable Reinforcement Learning: Methods, Targets and GapsResearch & Sciencecs.AI updates on arXiv.org8/10Relevance 8 out of 10
06:00CMASE: A multi-agent framework for real-time social intervention modelingResearch & Sciencecs.AI updates on arXiv.org8/10Relevance 8 out of 10