AI News

Learning How to Remember: Memory Modality Selection for GUI Agents (arxiv.org)

Beyond Discrimination: Evaluating Probability Calibration in Radiomics-Image Fusion for Renal Cell Carcinoma Classification (arxiv.org)

Beyond Gate Weights: Decision-Level Interpretation of Radiomics-Image Feature Fusion for RCC Classification (arxiv.org)

Next-Concept Modeling: Fast, Parallel Coarse-to-Fine Text Generation (arxiv.org)

Do you "see" what I "see"? A Multi-panel Visual Question Answer Dataset for Large Language Model Chart Analysis (arxiv.org)

Cortex: Measurement-Gated Evolution and Pruning of Coding-Agent Skills and Rules (arxiv.org)

Learnable Low-Rank Polynomial Sketch for Effective Linear Attention (arxiv.org)

CAPTOR: a vision-language foundation model for Chest-Abdomen-Pelvis Oncologic CT (arxiv.org)

Latent World Action Model for GUI Agents (arxiv.org)

Agentic Data Engineering for LLMs: System Design and a Controlled Empirical Study (arxiv.org)

Akule: Fast and Scalable Multi-Robot Motion Planning via Sparse Interaction Diffusion (arxiv.org)

GeoFix: Geometry Fixing for 4D Consistent Generation (arxiv.org)

RUBATO: Decoupled Clocks Allow Full-Duplex Models to Think as Fast as They Can (arxiv.org)

GameSmith: Scaling Zero-Sum Self-Play with Synthesised Games (arxiv.org)

EV-JEPA: Elastic V-JEPA for Any Device and Cross-Device Inference (arxiv.org)

TraceSampler: Making Sampling Errors Predictable and Correctable for Accelerated Generative Sampling (arxiv.org)

DexFLEX: Contact-Aware Foundation Controller for Command-Guided Dexterity (arxiv.org)

Deep learning from a control vantage point (arxiv.org)

D2: High-Capability Full-Duplex Spoken Language Models from Half-Duplex Backbones (arxiv.org)

PoLAR: Factorizing Extent and Mode in Latent Actions for Robot Policy Learning (arxiv.org)

X-Safe: Embodiment-Agnostic Safeguarding for Robot Arm Manipulation Policies via Action Masking (arxiv.org)

StableJudge: A Lightweight Adversarial Defense Framework for LLM-as-a-Judge (arxiv.org)

Growing to See: Efficient Pretraining of Self-Supervised Vision Transformers (arxiv.org)

DTA-GT: Direction- and Topology-Aware Graph Transformer for Neural Network Representation Learning (arxiv.org)

Unsupervised Data Drift Monitoring for Deployed Deepfake Speech Detectors (arxiv.org)

A Scaled-Up Empirical Study of Syntactic Language Models (arxiv.org)

Davinci-Tool: Scaling Agentic Tool-Use Supervision with Verifiable Executable Environments (arxiv.org)

Comparing GenAI to Machine Learning and Manual Review for eDiscovery (arxiv.org)

PrismQuant: Optimal Null-Space Rotations for Grouped Quantizers (arxiv.org)

Social Priors: Descriptions of AI Users Shape How Language Models Trust and Cooperate (arxiv.org)

RoboSimStudio: An Interactive Toolkit for Human and Agent Rapid Real2Sim of Robot Manipulation Tasks (arxiv.org)

KNN Implementation Details Can Dramatically Change Performance: An Example from Cover Trees (arxiv.org)

UA-Bench: Benchmarking Large Audio-Language Models on Underwater Acoustic Recognition (arxiv.org)

ProtoPur: Prototype-Purity Pseudo-Labeling for Weakly Supervised Instance Learning in Pathology (arxiv.org)

Privacy-Preserving GNN Inference on Financial Transaction Graphs via Homomorphic Encryption (arxiv.org)

ResearchFlow-Eval: Tracing the Execution Gap in End-to-End Multi-Claim Autoresearch (arxiv.org)

SarcasTraps: Effect of Sarcasm on Turkish Hate Speech Detection (arxiv.org)

Proactive Physical AI through Continuous Observation and Timely Intervention (arxiv.org)

The Value of Being Wrong: Self-Mined Visual In-Context Learning from Errors (arxiv.org)

WeGR: Unified Generative Retrieval for Industrial Search Systems in WeChat (arxiv.org)

From Suboptimal to Optimal: One-Step Latent Flow Policies for Offline Reinforcement Learning (arxiv.org)

AUGUstus: Autonomous Driving VLA Model with Unified Generation and Understanding (arxiv.org)

TabGift: A Context-Efficient Foundation Model for Tabular Data Generation (arxiv.org)

Governing Copyright Responsibly in the Era of Generative AI (arxiv.org)

MAGNET: Manifold-Aware Graph Diffusion Network for Connectome Generation (arxiv.org)

ProbabilityBench: Model Evaluation and Calibration with Exact Ground-Truth Uncertainty (arxiv.org)

CWE-bench: Refocusing Agentic Security from Exploitation to Defense (arxiv.org)

SPLICE: Sparse Physics Library Identification and Compositional Expansion (arxiv.org)

Gradient Routing Localizes and Removes Unintended Behaviors in RL (arxiv.org)

Superintelligence and the Triple Risk of an AI Race (arxiv.org)