DailyArxiv - AI Research Podcast

Today's papers: - End-to-end Optimization of Belief and Policy Learning in Shared Autonomy Paradigms: https://arxiv.org/abs/2601.23285v1 - Adaptive Edge Learning for Density-Aware Graph Generation: https://arxiv.org/abs/2601.23052v1 - OrLog: Resolving Complex Queries with LLMs and Probabilistic Reasoning: https://arxiv.org/abs/2601.23085v1 - From Abstract to Contextual: What LLMs Still Cannot Do i...

Today's papers: - VERSA: Verified Event Data Format for Reliable Soccer Analytics: https://arxiv.org/abs/2601.21981v1 - DynaWeb: Model-Based Reinforcement Learning of Web Agents: https://arxiv.org/abs/2601.22149v1 - AgenticSimLaw: A Juvenile Courtroom Multi-Agent Debate Simulation for Explainable High-Stakes Tabular Decision Making: https://arxiv.org/abs/2601.21936v1 - SymbXRL: Symbolic Explainabl...

Today's papers: - How do Visual Attributes Influence Web Agents? A Comprehensive Evaluation of User Interface Design Factors: https://arxiv.org/abs/2601.21961v1 - Vision-DeepResearch: Incentivizing DeepResearch Capability in Multimodal Large Language Models: https://arxiv.org/abs/2601.22060v1 - Retrieval-Infused Reasoning Sandbox: A Benchmark for Decoupling Retrieval and Reasoning Capabilities: ht...

Today's papers: - Investigating Associational Biases in Inter-Model Communication of Large Generative Models: https://arxiv.org/abs/2601.22093v1 - MoE-ACT: Improving Surgical Imitation Learning Policies through Supervised Mixture-of-Experts: https://arxiv.org/abs/2601.21971v1 - ToolWeaver: Weaving Collaborative Semantics for Scalable Tool Use in Large Language Models: https://arxiv.org/abs/2601.21...

Today's papers: - REASON: Accelerating Probabilistic Logical Reasoning for Scalable Neuro-Symbolic Intelligence: https://arxiv.org/abs/2601.20784v1 - A New Dataset and Framework for Robust Road Surface Classification via Camera-IMU Fusion: https://arxiv.org/abs/2601.20847v1 - CLEAR-Mamba:Towards Accurate, Adaptive and Trustworthy Multi-Sequence Ophthalmic Angiography Classification: https://arxiv....

Today's top AI research papers from arXiv: - Visual and Cognitive Demands of a Large Language Model-Powered In-vehicle Conversational Agent: https://arxiv.org/abs/2601.15034v1 - A Curriculum-Based Deep Reinforcement Learning Framework for the Electric Vehicle Routing Problem: https://arxiv.org/abs/2601.15038v1 - TIDAL: Temporally Interleaved Diffusion and Action Loop for High-Frequency VLA Control...

Today's papers: - Scalable Transit Delay Prediction at City Scale: A Systematic Approach with Multi-Resolution Feature Engineering and Deep Learning: https://arxiv.org/abs/2601.18521v1 - Assessing the Quality of Mental Health Support in LLM Responses through Multi-Attribute Human Evaluation: https://arxiv.org/abs/2601.18630v1 - Trust, Don't Trust, or Flip: Robust Preference-Based Reinforcement Lea...

Today's top AI research papers from arXiv: - Opportunities in AI/ML for the Rubin LSST Dark Energy Science Collaboration: https://arxiv.org/abs/2601.14235v1 - InT: Self-Proposed Interventions Enable Credit Assignment in LLM Reasoning: https://arxiv.org/abs/2601.14209v1 - "The Whole Is Greater Than the Sum of Its Parts": A Compatibility-Aware Multi-Teacher CoT Distillation Framework: https://arxiv....

An in-depth exploration of "The Assistant Axis: Situating and Stabilizing the Default Persona of Language Models" by Christina Lu and colleagues. This paper examines how language models maintain their helpful assistant persona and identifies a single geometric axis in activation space that controls persona behavior. The research has significant implications for AI safety and alignment, showing tha...

Today's top AI research papers from arXiv: - Clustering High-dimensional Data: Balancing Abstraction and Representation Tutorial at AAAI 2026: https://arxiv.org/abs/2601.11160v1 - Wetland mapping from sparse annotations with satellite image time series and temporal-aware segment anything model: https://arxiv.org/abs/2601.11400v1 - Beyond Model Scaling: Test-Time Intervention for Efficient Deep Rea...

Today's papers: - LongCat-Flash-Thinking-2601 Technical Report: https://arxiv.org/abs/2601.16725v1 - Standardizing Longitudinal Radiology Report Evaluation via Large Language Model Annotation: https://arxiv.org/abs/2601.16753v1 - Integrating Meteorological and Operational Data: A Novel Approach to Understanding Railway Delays in Finland: https://arxiv.org/abs/2601.16592v1 - Adoption of Generative...

Today's papers: - MMGRid: Navigating Temporal-aware and Cross-domain Generative Recommendation via Model Merging: https://arxiv.org/abs/2601.15930v1 - Cosmos Policy: Fine-Tuning Video Models for Visuomotor Control and Planning: https://arxiv.org/abs/2601.16163v1 - Pay (Cross) Attention to the Melody: Curriculum Masking for Single-Encoder Melodic Harmonization: https://arxiv.org/abs/2601.16150v1 -...

Today's top AI research papers from arXiv: - Cross-Modal Attention Network with Dual Graph Learning in Multimodal Recommendation: https://arxiv.org/abs/2601.11151v1 - Deep GraphRAG: A Balanced Approach to Hierarchical Retrieval and Adaptive Integration: https://arxiv.org/abs/2601.11144v1 - MetaboNet: The Largest Publicly Available Consolidated Dataset for Type 1 Diabetes Management: https://arxiv....

Today's top AI research papers from arXiv: - A Safety Report on GPT-5.2, Gemini 3 Pro, Qwen3-VL, Doubao 1.8, Grok 4.1 Fast, Nano Banana Pro, and Seedream 4.5: https://arxiv.org/abs/2601.10527v1 - Molmo2: Open Weights and Data for Vision-Language Models with Video Understanding and Grounding: https://arxiv.org/abs/2601.10611v1 - ProbFM: Probabilistic Time Series Foundation Model with Uncertainty De...

Today's top AI research papers from arXiv: - CogRail: Benchmarking VLMs in Cognitive Intrusion Perception for Intelligent Railway Transportation Systems: https://arxiv.org/abs/2601.09613v1 - LLM for Large-Scale Optimization Model Auto-Formulation: A Lightweight Few-Shot Learning Approach: https://arxiv.org/abs/2601.09635v1 - Full Disclosure, Less Trust? How the Level of Detail about AI Use in News...

Today's top AI research papers from arXiv: - SafeRedir: Prompt Embedding Redirection for Robust Unlearning in Image Generation Models: https://arxiv.org/abs/2601.08623v1 - WaterCopilot: An AI-Driven Virtual Assistant for Water Management: https://arxiv.org/abs/2601.08559v1 - Why AI Alignment Failure Is Structural: Learned Human Interaction Structures and AGI as an Endogenous Evolutionary Shock: ht...

Today's top AI research papers from arXiv: - Benchmarking Small Language Models and Small Reasoning Language Models on System Log Severity Classification: https://arxiv.org/abs/2601.07790v1 - Mon3tr: Monocular 3D Telepresence with Pre-built Gaussian Avatars as Amortization: https://arxiv.org/abs/2601.07518v1 - Puzzle it Out: Local-to-Global World Model for Offline Multi-Agent Reinforcement Learnin...

Today's top AI research papers from arXiv: - TowerMind: A Tower Defence Game Learning Environment and Benchmark for LLM as Agents: https://arxiv.org/abs/2601.05899v1 - Cedalion Tutorial: A Python-based framework for comprehensive analysis of multimodal fNIRS & DOT from the lab to the everyday world: https://arxiv.org/abs/2601.05923v1 - AdaFuse: Adaptive Ensemble Decoding with Test-Time Scaling for...

An in-depth exploration of "Learning Latent Action World Models In The Wild" - a paper that presents novel approaches to world modeling for reinforcement learning agents. We discuss how agents can learn to understand and predict the dynamics of complex environments without explicit action labels, the technical innovations behind this approach, and implications for building more capable AI systems...

Today's top AI research papers from arXiv: - GDPO: Group reward-Decoupled Normalization Policy Optimization for Multi-reward RL Optimization: https://arxiv.org/abs/2601.05242v1 - Observations and Remedies for Large Language Model Bias in Self-Consuming Performative Loop: https://arxiv.org/abs/2601.05184v1 - ConMax: Confidence-Maximizing Compression for Efficient Chain-of-Thought Reasoning: https:/...