AI News

From Local Evidence to Safety Verdicts: Causal Tracing in Vision-Language Models (arxiv.org)

Principles that Guide, Actions that Inform: Agent Evolution via Knowledge Abstraction (arxiv.org)

Metonymic Circuits for Abstract Concept Grounding in Vision Transformers (arxiv.org)

RadOnc-Agent: An LLM-Orchestrated Framework for AI Workflows Across the Radiotherapy Care Pathway (arxiv.org)

FluidPD: In-Place Elasticity for SLO-Aware Prefill-Decode Disaggregated LLM Serving (arxiv.org)

Learning to Remember: Distilling Memory Retention for Compact Recurrent Neural Networks (arxiv.org)

Evaluation of Active Feature Acquisition Policies with Tabular Foundation Models (arxiv.org)

Revisiting Temporal Regularization for Smooth Control in Deep Reinforcement Learning (arxiv.org)

ApexQuant: Data-Free Elastic Quantization by Residual Re-Isotropization (arxiv.org)

Stochastic Siamese MAE Pretraining for Longitudinal Medical Images (arxiv.org)

A Case Study in Assuring AI-Written Software (arxiv.org)

Text2Dashboard: A Governed Agent Architecture for Natural-Language Dashboard Generation over Enterprise DataBrain (arxiv.org)

GAMEGO: Training Game-Dev Agents with Synthetic Trajectories Anchored in Real-World Assets (arxiv.org)

OSFP4: Joint Optimization of Diagonal Smoothing and Block Scales for NVFP4 Quantization (arxiv.org)

UniST-Pred: A Robust Unified Framework for Spatio-Temporal Traffic Forecasting in Transportation Networks Under Disruptions (arxiv.org)

Machine learning Majorana topology using unsupervised and supervised learning (arxiv.org)

Recommender system in X inadvertently profiles ideological positions of users (arxiv.org)

Network Intervention by Polling Strategic Agents (arxiv.org)

Beyond Marginal Monitoring: Distributed Joint-Distribution Testing for Data Concept Drift in Large Scale E-Commerce Operations (arxiv.org)

Detecting a Shift Is Not Enough: Exact Minimax Limits of Linear Representation Repair (arxiv.org)

Spectra: Exact Component Transport for Test-Time Prior Adaptation in Simulation-Based Inference (arxiv.org)

Adaptive Mean Estimation by In-Context Learning: A Gradient-Flow Analysis (arxiv.org)

Extending Pathwise Gradients to Discrete Random Variables via Finite-Order Relaxation (arxiv.org)

Adversarially Trained Linear Transformers Are Optimal Robust In-Context Learners for Gaussian Mixtures (arxiv.org)

One-Shot Private Confidence Regions via Resampling (arxiv.org)

Exact Calibration and Sharp Risk Geometry for Volume-Sampled Ridge Regression (arxiv.org)

Asymptotic Analysis of Empirical Risk Minimization on Entry-wise i.i.d. Heavy-Tailed Data (arxiv.org)

Learning a Mixture of GFlowNets (arxiv.org)

Does Muon Need Fine-Grained Spectral Shaping? (arxiv.org)

The interface of data assimilation and machine learning (arxiv.org)

Structure, Not Belief: Correlated Thompson Sampling from LLM-Derived Covariance in Combinatorial Semi-Bandits (arxiv.org)

Reinforcement Learning for Hierarchical Reasoning Rewards: Minimax-Optimal Rates with Transformers (arxiv.org)

Active Feature Acquisition for Cost-Efficient Temporal Prediction with Reduced Participant Burden (arxiv.org)

Vine Copula VAR:From Recursive Margins to Joint Forecast Inference (arxiv.org)

A perspective note on likelihood approximation and inference for complex simulation models using a chain of aggregated normalizing flows (arxiv.org)

DeepAJM: Deep Association Joint Model for Irregularly Sampled data (arxiv.org)

Information-Dense Synthesis for Molecular Discovery (arxiv.org)

Redundancy and synergy in multivariate Gaussians via the Blackwell order (arxiv.org)

Empirical-Bayes spectral partial pooling across related tasks (arxiv.org)

Fine-Tuning VLM for Enhancing AI's Spatial Intelligence: Understanding 3D and 2D Rotations (arxiv.org)

Conditional Flow Matching for Transport Between Markov Processes (arxiv.org)

Poisson empirical Bayes estimation of sums of random variables via minimum-distance methods (arxiv.org)

A theory of platonic representations in language models (arxiv.org)

Sample-Optimal Estimation of the Fr\'echet Inception Distance (arxiv.org)

Multi-Task Active Learning with Efficient Resource Allocation (arxiv.org)

The Premise Is the Problem: Exchangeability Failure in Self-Monitored Test-Time Adaptation (arxiv.org)

Near-Optimal Sample Complexity for Recursive Entropic Risk Reinforcement Learning with a Generative Model (arxiv.org)

Anchor Divergence for Semantic Geometry in Contrastive Learning (arxiv.org)

Steering Diffusion Models to Rare Events with Sequential Monte Carlo (arxiv.org)

High-Dimensional Statistical Inference for Sparse Support Vector Machines (arxiv.org)