AI News

Predictable Failure in Multi-Hop Retrieval: Score-Distributional Confidence Scoring and Abstention (arxiv.org)

From Pretraining to Proficiency: Real-World Subtask RL for Long-Horizon Manipulation with Minimal Human Intervention (arxiv.org)

Gradient-Stable Attention Heads Signal LLM Correctness (arxiv.org)

CORDS: Continuous Representations of Discrete Structures (arxiv.org)

Rollout Total Correlation for Deep Reinforcement Learning (arxiv.org)

K2-V2: A 360-Open, Reasoning-Enhanced LLM (arxiv.org)

OverThink: Slowdown Attacks on Reasoning LLMs (arxiv.org)

Evolving Skill Modules under a Fixed Planner: Versioning, Rollback, and Runtime Governance for Long-Lived Robot Systems (arxiv.org)

PRIVET: PRoximIty leakage detection Via Extreme value Theory (arxiv.org)

Adaptive Rollout Truncation Based on Epistemic Uncertainty for Efficient Offline World Model Training (arxiv.org)

Purification and Regulation: Comorbidity-Aware Multi-Label Few-Shot Learning for Medical Image Classification (arxiv.org)

World Modeling in Transformers (arxiv.org)

GraphSkillEvo: Evolutionary Optimization of Graph-Structured Agent Skills (arxiv.org)

Calibrating Teacher--Student Discrepancy for On-Policy Distillation (arxiv.org)

RBS-Attention: Radius-Bounded Sparse Prefill for Long-Context Large Language Models (arxiv.org)

One Prompt Does Not Fit All: Self-Meta-Evolve for Personalized Information Extraction (arxiv.org)

Accelerating Dense LLMs via L0-regularized Mixture-of-Experts (arxiv.org)

GUARD: Natural Forgetting in Large Reasoning Models via Guided Answer-Reasoning Distillation (arxiv.org)

The Weight Is Over - Interactive Diffusion on Consumer GPUs (arxiv.org)

Neural Cellular Automata Learn General Features in their Hidden Channels (arxiv.org)

Detecting Pretraining Data in Large Language Models from a Free-Energy Perspective (arxiv.org)

LLMs as Feature Engineers for Text-and-Tabular Prediction (arxiv.org)

ExpBoN: Exponential-Noise Best-of-$n$ for Efficient Test-Time LLM Alignment (arxiv.org)

ECG Mirage: Revealing and Mitigating the Underutilisation of ECGs in Vision-Language Models for Clinical Prediction (arxiv.org)

LLM-Generated Feature Pools for Time Series Anomaly Detection (arxiv.org)

Matrix AdaGrad: Row-wise and Column-wise Adaptive Subgradient Methods (arxiv.org)

From Stress to Affect: Multimodal Deep Learning for Physiological Emotion Recognition Across Wearable Sensor Modalities (arxiv.org)

MOSAIC-SR: Transformer-Guided Symbolic Regression for Scientific Equation Recovery (arxiv.org)

On the Limits of Maximal Coding Rate Reduction for Out-of-Distribution Generalisation (arxiv.org)

Complex Problem Solving in Large Language Models: A Statistical Control Survey and Diagnostic Framework (arxiv.org)

Extreme classification: beating chance with one training example from each class (arxiv.org)

Bilevel Optimization of Topology and Hyperparameters (BOTH) (arxiv.org)

Reducing Barriers to Academic Support: Evaluating a Course-Specific RAG System for Addressing Help-Seeking Disparities in Higher Education (arxiv.org)

Beyond Accuracy: Centroid-Guided Contrastive Loss for Structured Fraudulent Job Posting Detection (arxiv.org)

Implicit Rule Induction with Test-Time Task Embeddings in ARC-like Tasks (arxiv.org)

SpecOpt: Contact-Diff Reasoning for Agentic Molecule Optimization Toward Binding Specificity (arxiv.org)

Can Agents Design Better Chips with a Higher Level Abstraction? (arxiv.org)

Clinician-Grounded Quality Assurance for AI-Assisted Psychiatric Intake (arxiv.org)

ASGARD: Action-Space Guard for UAV Resilience via Reinforcement Learning (arxiv.org)

A Smoothed Discrepancy Principle for Random Feature Methods and Neural Networks (arxiv.org)

AI-GRACE: A Use-Case Operationalization Framework for Agentic AI: From Organizational Objectives and Obligations to Deployment Capabilities and Architecture (arxiv.org)

Information-Gain Rewards over Diversity-Pruned Tests: GT-Anchored Verifier Co-Training for Reliable Code Generation (arxiv.org)

Triply-Scalable Equivariant Gaussian Process Modeling (arxiv.org)

Diagonalized Attention for Individualized Regression: Latent-Row Localization and Prediction (arxiv.org)

Ability-Residual Decoupled Modeling for Affective Cognitive Diagnosis (arxiv.org)

Sparse Identification for Automatic Large-Scale Screening: A Constraint-Aware Framework with Ultra Fast Decoding Algorithm (arxiv.org)

REFINEPPO: Learning Continuous Control Policies by Iterative Action Refinement (arxiv.org)

Do Quantum Models Scale Like LLMs? (arxiv.org)

Near-Optimal Acceleration for Smooth $\ell_p$ / $\ell_q$ Nondual Convex First-Order Oracle Optimization (arxiv.org)

GEM-MPC: Balancing Exploration and Exploitation through Expert-Guided Planning (arxiv.org)