AI News

Building Expressive and Tractable Probabilistic Generative Models: A Review (arxiv.org)

Delegation Without Trust: An Empirical Gap Analysis of Identity, Authorization, and Runtime Governance in Multi-Agent LLM Systems (arxiv.org)

CoLT-Drive: Counterfactual Long-Tail Benchmarking and Knowledge-Preserving Adaptation for Driving Affordance Prediction (arxiv.org)

Don't Let the Model Write the YAML: Deterministic, Minimal-Diff GitOps Remediation from LLM-Proposed Field Changes (arxiv.org)

Rock, Paper, Scissors, ... Dynamite - A Model of Disruption from New Technologies (arxiv.org)

Distributed Implicit Harm: A Compositional Safety Blind Spot in MLLM-Based Video Moderation (arxiv.org)

Assessing Suicide Risk in Arabic Crisis Helpline Calls: A Comparison of Arabic and English Large Language Models (arxiv.org)

Do General NLP Embeddings Capture Ontological Reasoning? (arxiv.org)

Scientific Agent Skills: A Library of Procedural Knowledge for Research Agents (arxiv.org)

Medical Causal Hypothesis Verification with Large Language Models (arxiv.org)

A Formal Analysis of Agent Payment Protocols (arxiv.org)

CUDA-Harness: Harnessing Agentic CUDA Kernel Generation and Optimization from Natural Language (arxiv.org)

Behaviorally Grounded User Profiles from the Wild for Personalized Alignment and Multi-Perspective Reasoning (arxiv.org)

InteractBench: Benchmarking LLMs on Competitive Programming under Unrevealed Information (arxiv.org)

Selective Agent Guidance via Entropy: Learning Autonomous Policies from Imperfect VLM Teachers (arxiv.org)

Can LLMs Discover Scientific Laws in Real and Parallel Worlds? (arxiv.org)

EvoSCM: Scientific Belief Revision Through Causal Model Evolution and Experimentation (arxiv.org)

When Guardrails Look Effective: Construct Validity Failures in LLM Agent Commerce Evaluation (arxiv.org)

Harness-of-Harness: Multi-Day Autonomous Software Development with Continual Improvement (arxiv.org)

Parsing the Stream: A Live Trace Model for Long-Horizon Agents and Their Observers (arxiv.org)

EdiTikZ: Scientific Figure Editing from Revision Trajectories (arxiv.org)

Neuro-Symbolic Geometric Abstraction (NeuSOGA): From Observations to Symbolic Mathematical Representations (arxiv.org)

SymFold: Synergizing Evolutionary and Structural Priors for Accurate Protein Inverse Folding (arxiv.org)

Process-Aware AI for Rainfall-Runoff Modeling: A Mass-Conserving Neural Framework with Hydrological Process Constraints (arxiv.org)

MMAI Gym for Science: Training Liquid Foundation Models for Drug Discovery (arxiv.org)

FloydNet: A Learning Paradigm for Global Relational Reasoning (arxiv.org)

Variance-Adaptive Muon: Pre-Orthogonalization Variance Modulation for Efficient Language Model Pretraining (arxiv.org)

Control Variate Score Matching for Diffusion Models (arxiv.org)

Training-Free Policy Violation Detection via Activation-Space Whitening in LLMs (arxiv.org)

Sierpi\'nski--Knopp Wasserstein Distance for Persistence Diagrams and Applications to 2-Wasserstein Approximation (arxiv.org)

Recent Developments in Transformer Inference Deployment on FPGA Platforms: A Survey (arxiv.org)

Does Imitation Learning Preserve Temporal Robustness in Dexterous Manipulation? An Expert-Learner Comparison Across Task Execution Speeds (arxiv.org)

Semantic-Guided Multimodal Preprocessing for Vision Transformer-Based Clear Cell Renal Cell Carcinoma Grading (arxiv.org)

Investigating Linear Probe Robustness to Linguistic Register, Medical Specialty, and Corpus Shifts in Medical QA (arxiv.org)

Exact Risk-Complexity Laws for Projective Boundaries in Scenario Optimization and Distribution-Free Certification (arxiv.org)

Where the Verifier Fails: A Category-Level Audit of Reward Signals in RLVR (arxiv.org)

Cheap Verifiers, Large Blind Spots: Measuring the Reliability Cost of Cost-Saving Cascades (arxiv.org)

GazeRefine: Expert Gaze as a Test-Time Prompt for Training-Free Medical Image Segmentation (arxiv.org)

Relational Task Generation Language: A Declarative Specification Framework for Relational Deep Learning (arxiv.org)

From Language to Behavior: Scaling Sequence Transformers for Industrial Recommendation Ranking with Rec-Native Designs (arxiv.org)

Accelerating Reinforcement Learning via MPC Solver-Gradient Guidance for Weights-varying MPC (arxiv.org)

AgentProv: Auditing Agentic LLM API Providers via Tool-use Policy Probes (arxiv.org)

Dense Weak Hiding: Closing Complexity Gaps in Nonconvex and PL Finite-Sum Optimization under Individual Smoothness (arxiv.org)

ES-AHD: An Evolution Strategy Framework for Automatic Heuristic Design (arxiv.org)

The Structure of Quantization Damage in LLMs: Why the Next Bit Should Be Spent Globally (arxiv.org)

How Do Language Models Choose Between Context and Memory? (arxiv.org)

Gradient-Update Mismatch: Rethinking Conflict-Free Training of Physics-Informed Neural Networks (arxiv.org)

CacheBridge: Efficient Cross-Model KV Cache Transfer (arxiv.org)

Denoising Diffusion Generative Models Secretly Calculate Attentions (arxiv.org)

Towards reliable multimodal disaster severity assessment through preference optimization and explainable vision-language reasoning (arxiv.org)