AI News

Physical AI Governance: From Theory to Practice Across Life Cycle (arxiv.org)

When Every Simulation Counts: Value-Based Reinforcement Learning for Accelerated Photonics Inverse Design (arxiv.org)

Limbomorphs (arxiv.org)

Lions and Muons: Optimization via Stochastic Frank-Wolfe under Heavy-Tailed Noise (arxiv.org)

xMIx: High-Performance Serving-Time Platform for Mechanistic Interpretability Apps (arxiv.org)

Beyond Block Boundaries: Multi-Block Editing for Diffusion Large Language Models (arxiv.org)

Structure Over Scale: Schema-Constrained Causal Graphs for RAG (arxiv.org)

Differencing the Diffusion Trajectory toward Uncertain Components for Time Series Forecasting (arxiv.org)

Opti-Q: A Constraint-Based Optimization Framework for Multi-LLM Question Planning (arxiv.org)

ACM: Agentic Context Management for Long Horizon Tasks (arxiv.org)

HyCE-RAG: Hypergraph Chain-of-Evidence Retrieval-Augmented Generation for Explainable Multi-hop Question Answering (arxiv.org)

Adaptive Data Admission and Retention for Streaming Federated Learning (arxiv.org)

Constrained Reinforcement Learning Using Successor Representations (arxiv.org)

Capacity-Aware Deep Learning for Generalizable Traffic Volume Estimation Across Links and Cities (arxiv.org)

When Should Active RAG Retrieve? A Budget-Aware Evaluation of Utility, Calibration, and Cost (arxiv.org)

Distributional Random Forests for Complex Survey Designs (arxiv.org)

DynaCalKV: Key-Value Cache Compression via Head Grouping and Adaptive Rank Allocation (arxiv.org)

Agent Team Work Zone: An Automated, Persistent Workspace for Long-Lived Coding Agent Teams (arxiv.org)

Chart Deception in Vision-Language Models: From Vulnerability to Mitigation (arxiv.org)

K-Survival Means (arxiv.org)

Hierarchical Group-Conditional Conformal Risk Control for Selective Prediction in Language Models (arxiv.org)

MAPLE: Efficient and Diverse Multi-Alpha Generation for Portfolio Construction (arxiv.org)

Learning Reusable Hybrid Motion Priors for Humanoid Locomotion from Motion Imitation (arxiv.org)

LLM-Assisted Ontology Engineering and Construction of a French Legal Knowledge Graph (arxiv.org)

Stacking the Deck: Tunable Trainability in Stacked LCUs (arxiv.org)

An Empirical Study of Feature Selection Granularity (arxiv.org)

How Well Can AI Generate Backlogs from App Mockups? (arxiv.org)

Self-Motivated Growing Neural Network for Adaptive Architecture via Local Structural Plasticity (arxiv.org)

Coordinated Networking for On-Device Agent-Augmented Real-Time Communication (arxiv.org)

Evaluating LLMs as Interpretable Controllers for Dynamical Systems (arxiv.org)

From RLVR to RLSVR: Task Transformation Induces Self-Verifiable Rewards for Open-Ended LLM Self-Improvement (arxiv.org)

MANGO: A Global Single-Date Paired Dataset for Mangrove Segmentation (arxiv.org)

The Intruder Threshold: A Spectral Law for LoRA Fine-Tuning (arxiv.org)

A Statistical Difference between Single-Layer Learning and Hierarchical Learning in Wide Neural Networks (arxiv.org)

Decentralized Granular Access Control for Agentic AI Systems in Critical Infrastructure (arxiv.org)

LLM-SoccerArena: Benchmarking LLMs on Real-World Predictions in Sports (arxiv.org)

Asymmetric Hierarchical Anchoring for Robust Audio-Visual Cross-Modal Generalization (arxiv.org)

Learning Distributions from Multiple Data Providers (arxiv.org)

Learning to Access Computation: Accessibility Plasticity as a Principle of Adaptive Intelligence (arxiv.org)

Forgetting is Everywhere (arxiv.org)

ClinFusion: A Vision-Centric Multimodal LLM System for Holistic Medical Understanding (arxiv.org)

Attenuated Heterogeneity in Fixed-Effects Causal Forests, and a Cross-Fitted Correction (arxiv.org)

Every Client Is an Environment: Federated De-confounding for Spatio-Temporal Forecasting (arxiv.org)

Masked Autoencoders Learn Perception-Relevant Representations from Resting State Neural Data (arxiv.org)

From Hybrid Mechanistic--Data-Driven Modeling Toward Neuro-Symbolic AI: What, Why, and How (arxiv.org)

ATLAS: Automated Approximation of Transformers for Efficient Homomorphic Inference in One Hour (arxiv.org)

Monitoring Post-Disaster Urban Recovery Using High-Resolution SAR Time Series and Unsupervised Learning: Evidence from the 2023 T\"urkiye-Syria Earthquake (arxiv.org)

Semalith v1.4: A Calibrated 184M Safety Classifier Achieving State-of-the-Art Prompt-Injection Detection at 44x Fewer Parameters than Llama-Guard-3-8B (arxiv.org)

Efficient Learning of Truncated Boolean Product Distributions: Influence to the Rescue (arxiv.org)

PathRIR: Physics-Guided Acoustic Path Selection and Late-Tail Compensation for Fast Room Impulse Response Simulation (arxiv.org)