AI News

Authority-Preserving Evaluation of Medical Vision-Language Assistants (arxiv.org)

Listen Then Reason: Perception-Grounded Test-Time Reinforcement Learning for Large Audio-Language Models (arxiv.org)

Pinocchio: Fast Uncertainty Estimates for Black-Box Language Models (arxiv.org)

A Global Comparison of Schemas, Transparency, and Interoperability in Public-Sector AI Registers and Inventories (arxiv.org)

BackTrend: Evaluating Scientific Weak-Signal Prediction via Backward Reconstruction (arxiv.org)

Multi-Term Fourier Graph Neural Network with Sample Relationship Learning for Enhanced Remaining Useful Life Prediction (arxiv.org)

On The Robustness-Resolution Tradeoff In Temporal Quantization Of Event Streams (arxiv.org)

Validating, Not Sampling: Region-Level Robustness of Vision-Language and Vision-Language-Action Models (arxiv.org)

Do Chess Explanations Reflect Model Decisions? Behavioral and Token-Level Tests of LLM Reasoning Faithfulness (arxiv.org)

ORDER: A Fictitious-World Benchmark for Domain-Adaptive Embodied AI (arxiv.org)

Sample-Smooth Spaces: A Convenient Category for Differentiable Probabilistic Programming (arxiv.org)

Quantifying Protocol-Induced Uncertainty in Comparative Predictive-Model Evaluation: Evidence from Large-Scale Daily PM10 Forecasting (arxiv.org)

CompKV: Compensation-Aware KV Selection for Long-Context LLM Inference (arxiv.org)

PreGS: A Parameter-Transfer-Based Multi-Expert Graph Neural Network for Node Classification (arxiv.org)

Disaggregated Quantization: Specializing LLM Prefill and Decode (arxiv.org)

On the Effect of Bit-Level Parameter Perturbations in Machine Learning and Deep Learning Models (arxiv.org)

SyzHarness: Patch-Based Kernel Bug Reproduction with LLM-Synthesized Fuzzing Harnesses (arxiv.org)

PACT: From Credit Assignment to Critic Alignment (arxiv.org)

FairMean: Promoting Fairness in Distributed Learning under Label Poisoning Attacks (arxiv.org)

Learning to Defer with Guidance on Real World Medical Data (arxiv.org)

TimeInteract: Towards Real-Time Interactive Intelligence for Streaming Time Series (arxiv.org)

Double Descent and Malign Overfitting in Diffusion Models (arxiv.org)

Complementary rPPG-Derived and Lip-Region Frequency Cues for Talking-Face Deepfake Detection (arxiv.org)

Rethinking Streaming Video Diffusion Model: Context, Execution, and Training (arxiv.org)

Brain-to-Image Generation: Reconstructing Visual Stimuli from EEG using Generative Adversarial Networks (arxiv.org)

OMatG-flash: An All-Atom Flow Map with Reinforce Adjoint Matching for Scalable Materials Discovery (arxiv.org)

RoofLang: Enabling AI-Driven Architecting of LLM Inference Systems (arxiv.org)

DeepFEAv2: Deep Learning for Transient Finite Element Analysis Beyond Structured Meshes (arxiv.org)

Can We Predict Anomaly Detection Performance from Embedding-Space Geometry? (arxiv.org)

Performance vs Consistency: Evaluating a Foundation Model in Lung-RADS Screening (arxiv.org)

Glucose-ML: A collection of longitudinal diabetes datasets for development of robust AI solutions (arxiv.org)

When Recursive Models Finish Computing (arxiv.org)

Notes on Fourier-Bessel wavelets (arxiv.org)

Towards Hierarchical GNNs for multi-grid power flow: generalization across operating scenarios (arxiv.org)

TransBERT: A Framework for Synthetic Translation in Domain-Specific Language Modeling (arxiv.org)

Greedy Decoding Is Not Precision-Invariant: Cross-Precision Output Divergence in LLM Inference (arxiv.org)

Riemannian Optimization on Tree Tensor Networks with Application in Machine Learning (arxiv.org)

Exact and Approximate Range Queries in Ball Mapper (arxiv.org)

Label-Efficient Learning for Ground-Based Sky-Image Classification: A Benchmark of Transfer Learning, Active Learning, and Pseudo-Labeling on GCD (arxiv.org)

MAGIC: Mixed-Granularity Agent Graphs via Incremental Construction with Dense-Reward Reinforcement Learning (arxiv.org)

A Spectral Theory of Grokking: Weight Decay induces Feature Learning (arxiv.org)

The Self Driving Portfolio: Agentic Architecture for Institutional Asset Management (arxiv.org)

Train Where the Quantized Model Goes: On-Policy Distillation for Low-Bit Reasoning (arxiv.org)

The Sirens' Song: When Proximal Background Context Overshadows Distant Evidence (arxiv.org)

EquivSVA: A Formally Verified Dataset of Behavioral Assertions Across Equivalent RTL Implementations (arxiv.org)

From IceCube to IT-Sphere: A Hybrid Quantum-Classical GNN for Banking IT Root Cause Analysis (arxiv.org)

Target alignment, dilution and forecast selection when cross-sectional forecasts share a common target (arxiv.org)

A Survey of Agentic Reasoning for Large Language Models: Towards Recursively Self-Improving and Collective Agents (arxiv.org)

VeRA: Renewing Reasoning Benchmarks with Executable Specifications (arxiv.org)

When Scaling Fails: Mitigating Audio Perception Decay of LALMs via Multi-Step Perception-Aware Reasoning (arxiv.org)