AI News

RT-Safe: Benchmarking Agent Safety in Real-Time Embodied Environment (arxiv.org)

CATune: Structural Constraint-Aware Bayesian Optimization for DBMS Configuration Tuning (arxiv.org)

An Empirical Study of Agent Skills' Downstream Utility (arxiv.org)

Learned Monotone Recurrent Features in Governed Credit Scoring: The Price of the Frame and the Necessity of Macro Conditioning (arxiv.org)

Geometry-Aware Diffusion Approximate Posterior Sampling for Sparse-View and Limited-Angle CT (arxiv.org)

Adversarial RL for Port-Scan Evasion: Attacker Feature Visibility in Edge-Deployed IDS (arxiv.org)

LRCC: Generalizing Low-Rank Compression with Conditional Computation (arxiv.org)

Autonomous Droplet Navigation via Model-Based Reinforcement Learning: Zero-Shot Transfer and Emergent Dynamics (arxiv.org)

Beyond Risk Prediction: Evidence Grounding and Psychosocial Factor Verification for Explainable Suicide Risk Assessment (arxiv.org)

Beyond the Sycophancy Score: How Task, Model, and Pressure Shape LLM Yielding (arxiv.org)

CoDR: Training-Free Confidence-Drift Remasking for Diffusion Language Models (arxiv.org)

Child ASR Adaptation with Adult Retention: An Empirical Study (arxiv.org)

Route-Verify-Vote: Procedure-Conditioned Self-Consistency for Mixed-Domain Reasoning (arxiv.org)

On the Cyclic Assumption of the Cow-Path Search Algorithm (arxiv.org)

Accelerating Floating-Point Satisfiability Solving via Gradient Normalization (arxiv.org)

HE-OFT: Privacy-Preserving One-Shot Federated Fine-Tuning under Homomorphic Encryption (arxiv.org)

Deep Learning-Based Tri-Hybrid Multi-User MIMO Precoding: The Blessing of EM-Reconfigurable Antennas (arxiv.org)

Decoupling Exploration from Optimization in RLVR (arxiv.org)

Distilling Graph Geometry: Knowledge Gap from GNNs to MLPs (arxiv.org)

Composing What Each Teacher Learned: Multi-Teacher On-Policy Distillation through Teacher-Relative Shifts (arxiv.org)

NeuralBES: A Differentiable, Control-Aware Emulator for Scalable Building Energy Modeling (arxiv.org)

A Good Self-Teacher Meets the Student Where They Are: Joint On-Policy Learning and Teaching (arxiv.org)

Seq-Flow: Efficient Probabilistic Forecasting with Self-Rollout Error Control (arxiv.org)

Q-Learning with Scalar Adjoint Matching (arxiv.org)

Which Rollout Taught It That? BehaviorTrace and the Limits of Training-Data Attribution in Online RL (arxiv.org)

From Plausible Hierarchies to Useful Taxonomies: Evaluating Agentic Harnesses on Customer Feedback (arxiv.org)

OnlineQAT: On-Policy Distillation for Ultra-Low-Bit Large Language Models (arxiv.org)

Cross-Domain Pretraining for Steady-State Neural CFD Surrogates (arxiv.org)

Executing Causal Structure Learning with Linear-Attention Transformers (arxiv.org)

Matching of signal, noise and hardware timescales for filtering and forecasting of correlated noise signals (arxiv.org)

Kernel Autoresearch for Open-Ended Model Discovery (arxiv.org)

ResidualQuant: KV Cache Quantization for Looped Transformers with 2-Bit Residuals (arxiv.org)

Continual Learning without Continual Training (arxiv.org)

Input-Blind Controls Produce Substantial Oracle Headroom for Layer Programs in Multiple-Choice Evaluation (arxiv.org)

Temporally Interpretable Differentiable Decision Trees (arxiv.org)

ORDERS: An Empirical Study of Norm-Rank Aggregation for Personalized Federated Learning (arxiv.org)

Average-Reward Reinforcement Learning for Multichain MDPs: A Hierarchical Decomposition Approach (arxiv.org)

HAN-Mamba: Hierarchical Selective State Space Networks for Multi-Scale Financial Volatility Forecasting (arxiv.org)

Thinking in Depth: Retrospective Inference for Tabular Foundation Models (arxiv.org)

PoreML: A Data-Driven Framework for Learning Multiphase Flow in Porous Media (arxiv.org)

Fault-tolerant foundation models (arxiv.org)

RSIGym: A Flexible Environment for Recursive Self-Improvement (arxiv.org)

How Do Transformers Learn to Represent Symmetries? (arxiv.org)

SemanticFold: Latent Sequence Compression SeparatesLanguage Modeling, Decodability, and Reasoning (arxiv.org)

Continual Graph Multi-Agent Reinforcement Learning (arxiv.org)

Ask the Expert: LLM-Guided Reinforcement Learning for Autonomous Cyber Defense (arxiv.org)

Physics-Aligned Electronic Ground-State Learning Improves Generalization (arxiv.org)

PatchBench: Measuring Collateral Damage in Activation Patching (arxiv.org)

Sparse Planning in Visual World Models via Cost Gradients (arxiv.org)

A Closed-Loop Non-Asymptotic Convergence Analysis of PPO with Learned Critics and Clipping (arxiv.org)