AI News

DIP: Dynamic In-Context Planner For Diffusion Language Models (arxiv.org)

Entropy-Aware Token Rejection for Improving Speculative Decoding (arxiv.org)

SGM: Safety Glasses for Multimodal Large Language Models via Neuron-Level Detoxification (arxiv.org)

CoFiRec: Coarse-to-Fine Tokenization for Generative Recommendation (arxiv.org)

WaymoQA: A Multi-View Visual Question Answering Dataset for Safety-Critical Reasoning in Autonomous Driving (arxiv.org)

On the Optimality of Kinship Naming: an Information-theoretic Approach (arxiv.org)

Look It Up: Analysing Internal Web Search Capabilities of Modern LLMs (arxiv.org)

Error-Driven Scene Editing for 3D Grounding in Large Language Models (arxiv.org)

Adversarial Trust Poisoning in Vehicular Collaborative Perception (arxiv.org)

Adaptive Multi-Branching for Shallow Decision Tree Induction (arxiv.org)

When Do Larger Batches Help Scale LLM Reinforcement Learning? (arxiv.org)

A Spectral Identifiability Threshold for Dissipative Rate Recovery from Truncated Liouvillian Spectra (arxiv.org)

MEL: Coordinate-Preserving EEG Tokenization for fMRI Translation (arxiv.org)

GFlowNets and variational inference (arxiv.org)

Uncertainty-Aware Multi-Task Learning for Joint Modulation Recognition and SINR Estimation (arxiv.org)

Information-Based Calibration of Uncertainty Quantification in Product-of-Experts Gaussian Process Models (arxiv.org)

Spatial Entropy based Partitioning for Spatiotemporal Graph Unlearning (arxiv.org)

Unlearning on Spatio-Temporal Graphs through Subgraph Virtual Edge Reconstruction (arxiv.org)

Fully Distributed GNE Algorithms for Multi-Robot Placement without Consensus on Multipliers (arxiv.org)

ProgRouter: Online Progress-Guided Orchestration for Multi-Agent LLM Workflows under Quality-Cost Tradeoffs (arxiv.org)

Candidate supply and answer selection shape the value of LLM judging in multi-agent systems (arxiv.org)

Repair or Resample? Rethinking Failure Debugging in LLM Multi-Agent Systems (arxiv.org)

Where Induction Runs Out: Description-Length Difficulty and the Memorisation Gap in Integer-Sequence Benchmarks (arxiv.org)

Scalable Clinical Data Infrastructure and Comparative ML Evaluation for Hospitalisation Risk Prediction in Elderly Patients with Multiple Long-Term Conditions using CPRD (arxiv.org)

One Capability or Many? Testing the Economic Validity of Frontier AI Evaluation (arxiv.org)

Behavioral Latency as Weak Event-Time Supervision for EEG Reaction-Time Decoding (arxiv.org)

Does Latent Planning Survive Point Clouds? Action-Conditioned JEPA World Models for Geometric Observations (arxiv.org)

post-graph-rag: A PostgreSQL-Native Bi-Temporal Graph RAG Engine with Temporal Grounding at Synthesis (arxiv.org)

SimGuide: Typed Multi-Context User Representations for Preference-Conditioned Agent Planning (arxiv.org)

Reference-Grafting Matches Fine-Tuning at Eliciting Sandbagged Capabilities (arxiv.org)

A Causal Model for Locating and Unlocking Sandbagging in Model Organisms (arxiv.org)

Knowledge Distillation under Teacher Misspecification: An Order-Parameter Analysis of the Gap between Teacher Mimicry and Task Performance (arxiv.org)

Learning Human Health and Diseases from 24-hour Wrist Movement (arxiv.org)

Exponential random graph models with soft clique constraints (arxiv.org)

HumanStudy-Bench: Towards AI Agent Design for Participant Simulation (arxiv.org)

Target-Aware State-Adaptive $p$-Dirichlet Graph Neural Regression for Non-Invasive Body-Composition Estimation (arxiv.org)

Adversarial Online Classification with a Preview (arxiv.org)

Denoising as Projection: Constrained Optimization with Gradient-Guided Diffusion (arxiv.org)

On the Plasticity Collapse in Continual Machine Unlearning (arxiv.org)

MedCache: Efficient and Temporally Valid Memory for Longitudinal Clinical Agents (arxiv.org)

BEACON: Behavioral and Semantic Enrichment of AlphaEarth Embeddings through Tri-Modal Contrastive Learning (arxiv.org)

Asynchronous Cooperative Online Learning for Multi-Robot Control under Computational Delays (arxiv.org)

HoopMind: A Real-Time Neural Game-Tree System for Opponent-Aware Possession Planning (arxiv.org)

Event-triggered Control and Online Learning for Networked Systems under Computational Delays (arxiv.org)

Simulation-Based Evaluation of Energy-Constrained Quantum-Classical Competition (arxiv.org)

Predicting the Unpredictable: LLM-powered Long-term Chaotic Time Series Forecasting under Short-term Observations (arxiv.org)

On the Resilience of Text-to-Video Diffusion Models to Hardware Faults (arxiv.org)

A Multi-Agent Human-LLM Collaborative Framework for Closed-Loop Scientific Literature Summarization (arxiv.org)

DART: Draft-Agreement Routing for Training-Free Adaptive Thinking Budgets in Hybrid Reasoning Models (arxiv.org)

Delta-AI: Local objectives for amortized inference in sparse graphical models (arxiv.org)