AI News

Evaluating GNNs for Success Prediction in Artist Collaboration Networks (arxiv.org)

Judging LLM-as-a-Judge: Concerning Rubric Artifacts in LLM-based Automated Text Generation Evaluation (arxiv.org)

BharatGather: A Culturally-Informed Benchmark Dataset for Misinformation and Fake News Detection in Indian Public Events (arxiv.org)

Robust PAC Learning of Concurrent Stochastic Games (arxiv.org)

A Low-Cost, Open Platform for End-to-End Autonomous Driving on a Miniature Ackermann Vehicle (arxiv.org)

Prospective Coding Improves Learning in Deep Continuous-Time Recurrent Networks (arxiv.org)

Listen to the Latents: Self-Correcting Speech Recognition in Large Audio Language Models Through Hidden-State Interactions (arxiv.org)

Constant regret in general games via higher-order optimism (arxiv.org)

Hardware-Aware FP4 FlashAttention-4 (arxiv.org)

Conditioning Degenerate Diffusion Models (arxiv.org)

Subspace Inference Enables Efficient Active Reward Learning from Preferences (arxiv.org)

A location-invariant estimator of extremal quantile treatment effects for heavy-tailed distributions (arxiv.org)

Unlocking Lossless Speedups in LLMs via Discrete Diffusion (arxiv.org)

RobustSeiz: An Open-Source Framework for Benchmarking the Robustness of EEG Seizure Detection Models (arxiv.org)

OSR: Output Space Redistribution for Adaptive Label Removal in Classification Models (arxiv.org)

VestigeKV: The NoPE-MLA KV Cache Carries Its Own Eviction Signal in a Vestigial Branch (arxiv.org)

From Deceptive Outputs to Deceptive Mechanisms: A Causal Framework for Language-Model Deception Research (arxiv.org)

Headroom-Drift Replay: A Primitive for Principled Replay Control in GRPO (arxiv.org)

RATL: Learning from Retrieved Residuals for Robust Multivariate Time-Series Forecasting (arxiv.org)

Beyond Endpoint Scores: Time- and Capacity-Conditioned Evaluation of Continual Knowledge Updating (arxiv.org)

Differentiable Interval Bottlenecks for Interpretable Anomaly Detection in Numerical Data (arxiv.org)

Pushing the (Decision) Boundaries: Dynamically Calibrating Differentially Private Noise to Explainability in Federated Learning (arxiv.org)

Multi-step Proximal Policy Improvement in Offline Reinforcement Learning (arxiv.org)

Witnesses Explain Anomalies (arxiv.org)

A Peer-Relative Representation Learning Framework for Energy Inefficiency Identification in Mobile Network Sites (arxiv.org)

Landmark-Based Discrimination of Injury-Associated Athlete-Sessions from Minute-Resolution Multimodal Football Monitoring Data (arxiv.org)

OBER+: Continuity-Aware Reporting and Traceable Continuous Improvement in Outcome-Based Education (arxiv.org)

From Nowcasting to Forecasting: Adapting a Reanalysis-Trained (arxiv.org)

Projected Riemannian Gradient Descent for the Bures-Wasserstein Barycenter: Dimension-Independent Linear Convergence at Unit Step Size (arxiv.org)

Federated Causal Discovery via Regression-Directed Cumulants (arxiv.org)

Resolution-Aware Experimental Design under Partial Identifiability (arxiv.org)

Extracting Forgotten Prompts from Targeted Unlearned Models (arxiv.org)

Local Updates, Global Learning (LUGL): Playing Games with non-incremental Learners (arxiv.org)

Neural-Network Maxent: a general extension with learned nonlinearity, applied to time-series for Desert Locust distribution modelling (arxiv.org)

WeatherNext 3: Increasing resolution and performance of global weather models with raw observations (arxiv.org)

One Editor, Many Edits: A Unified Training-Free Framework for Diverse Video Editing (arxiv.org)

Interface-Induced Trajectory Censoring (arxiv.org)

FiMI Banking: A Sovereign Model for Indian Retail Banking (arxiv.org)

More Criticism Does Not Make a Better Review: EquiReview-R (arxiv.org)

Towards Numerical TOHTN Planning with SMT-based HTN-SAT Encoding (arxiv.org)

Speak for Me: Giving LLMs the Situational Awareness to Participate in a Meeting (arxiv.org)

Common-Witness Certificates and Sharp Feature Bounds for Counterfactual Image Auditing (arxiv.org)

Value-Preserving Architectures for Agentic AI Systems (arxiv.org)

Lose the Order, Keep the Hierarchy: Deordering HTN Plans (arxiv.org)

Inferring Affective Consciousness in an Artificial Agent: A Case Study (arxiv.org)

Xiaomi-TabLDM: A Tabular Foundation Model Technical Report (arxiv.org)

STAIR (STructure Aware Information Retriever): A novel dataset and LLM based retriever for document structure augmentation (arxiv.org)

Bioinfoysis Technical Report (arxiv.org)

Adapting to Evolving Requirements: Agentic AI for Retail Supply Chain Operations (arxiv.org)

Semantic Bayesian World Models (arxiv.org)