AI News

MOSAIC: Efficient Mixture-of-Agent Scheduling via Adaptive Aggregation and Inference Concurrency (arxiv.org)

ThoughtFold: Folding Reasoning Chains via Introspective Preference Learning (arxiv.org)

What Do Students Learn? A Feature-Level Analysis of Dark Knowledge (arxiv.org)

Overlaying Governance: A Compositional Authorization Framework for Delegation and Scope in Agentic AI (arxiv.org)

How Quantization Changes Interpretable Features: A Sparse Autoencoder Analysis of Language Models (arxiv.org)

CoughSense: Five-Class Respiratory Disease Classification via Whisper Encoder Fine-Tuning and Dual-Encoder Cross-Attention Fusion with Balanced Contrastive Learning (arxiv.org)

From Prompt to Service: An SLM-Based Agent Orchestration Gateway for AI-Driven Virtual Worlds (arxiv.org)

Assistax: A Multi-Agent Hardware-Accelerated Reinforcement Learning Benchmark for Assistive Robotics (arxiv.org)

Localized, High-resolution Geographic Representations with Slepian Functions (arxiv.org)

Exact equivariance, kept through training, buys zero-shot generalisation across the symmetry group (arxiv.org)

Hierarchical RBF-KAN and RBF-SKAN Architectures for Multidimensional Function Approximation and Random Field Learning (arxiv.org)

ERP-XTTN: Interpretable Prototype-Guided Cross-Attention for Cross-Subject ERP Classification (arxiv.org)

InfoMem: Training Long-Context Memory Agents with Answer-Conditioned Information Gain (arxiv.org)

BYORn: Bootstrap Your Own Responses to Defend Large Vision-Language Models Against Backdoor Attacks (arxiv.org)

Synthesize and Reward -- Reinforcement Learning for Multi-Step Tool Use in Live Environments (arxiv.org)

A formal definition and meta-model for a machine theory of mind (arxiv.org)

Efficient Hyperparameter Optimization for LLM Reinforcement Learning (arxiv.org)

Fast Unlearning at Scale via Margin Self-Correction (arxiv.org)

Gate AI: LLM Security Benchmark Evaluation Methodology and Results (arxiv.org)

Perceive Before Reasoning: A Pre-Reasoning Perception Framework for Efficient and Reliable Proactive Mobile Agents (arxiv.org)

BAHSD: Bridging the Long-tail Gap via Adaptive Distillation in Black-box Sequential Recommendation (arxiv.org)

PhotoCraft: Agentic Reasoning with Hierarchical Self-Evolving Memory for Deep Image Search (arxiv.org)

AnyAudio-Judge: A Dynamic Rubric-Based Benchmark and Evaluator for Audio Instruction Following (arxiv.org)

Spike-Aware C++ INT8 Inference for Sparse Spiking Language Models on Commodity CPUs (arxiv.org)

Fully Automated Identification of Lexical Alignment and Preference-Stage Shifts in Large Language Models (arxiv.org)

Trading Human Curation for Synthetic Augmentation in RLVR (arxiv.org)

Multi-Segment Attention: Enabling Efficient KV-Cache Management for Faster Large Language Model Serving (arxiv.org)

Coherent Swap Regret and Channel-Proof Learning (arxiv.org)

Staying Alive: Uncensored Survival Analysis with Tabular Foundation Models (arxiv.org)

When Graph Tokens Sink: A Mechanistic Analysis of Graph Language Models (arxiv.org)

AVTrack: Audio-Visual Tracking in Human-centric Complex Scenes (arxiv.org)

EntangleCodec: A Unified Discrete Audio Tokenizer via Semantic-Acoustic Entanglement (arxiv.org)

NVIDIA OmniDreams: Real-Time Generative World Model for Closed-Loop Autonomous Vehicle Simulation (arxiv.org)

Decoupled Smart Contract Audits: Lightweight LLM Framework via Distillation and Aggregation (arxiv.org)

How Many Trees in a Random Forest? A Revisited Approach with Plateau Search and Optuna Integration (arxiv.org)

Training a Predictive Coding Network on ImageNet using Equilibrium Propagation (arxiv.org)

Hybrid Adaptive Kalman Filtering for Data-Efficient Joint Tracking and Classification (arxiv.org)

CauTion: Knowing When to Trust LLMs for Ensemble Causal Discovery (arxiv.org)

AI Rater Discrimination Depends on Scoring Protocol in Complex Clinical Decision-Making (arxiv.org)

The Ringelmann Effect in Multi-Agent LLM Systems: A Scaling Law for Effective Team Size (arxiv.org)

Learning to Refine: Spectral-Decoupled Iterative Refinement Framework for Precipitation Nowcasting (arxiv.org)

Inference Cost Attacks for Retrieval-Augmented Large Language Models (arxiv.org)

Are Common Substructures Transferable? Riemannian Graph Foundation Model with Neural Vector Bundles (arxiv.org)

Post-Hoc Robustness for Model-Based Reinforcement Learning (arxiv.org)

Demystifying Pipeline Parallelism: First Theory for PipeDream (arxiv.org)

A Geometric Lens on Physics-Aligned Data Compression (arxiv.org)

SVHalluc: Benchmarking Speech-Vision Hallucination in Audio-Visual Large Language Models (arxiv.org)

When RLHF Fails: A Mechanistic Taxonomy of Reward Hacking, Collapse, and Evaluator Gaming (arxiv.org)

EqGINO: Equivariant Geometry-Informed Fourier Neural Operators for 3D PDEs (arxiv.org)

Whom to Query for What: Adaptive Group Elicitation via Multi-Turn LLM Interactions (arxiv.org)