AI News

GeoDose-CP: Graph-Local Conformal Inference for Continuous-Treatment Earth Observation (arxiv.org)

Beneath the Scores: Rethinking Hallucination Evaluation for Video Understanding Models (arxiv.org)

CrossSafe: Towards Cross-Embodiment Latent Safety Filters (arxiv.org)

Cross-Country Code-Mixing for Generative Recommendation (arxiv.org)

Persuaded, Not Informed: Incentive-Misaligned Witnesses Defeat In-Context Grounding (arxiv.org)

Blockchain-Enabled Artificial Intelligence and AI Agents for Secure Data Sharing and Cybersecurity Applications (arxiv.org)

Forecast-Dojo: Replayable Environments for Benchmarking and Training LLM Forecasting Agents (arxiv.org)

When Fancy Eviction Fails: Rethinking Cache Replacement For LLM Prefix Reuse (arxiv.org)

Multimodal Routing and Region Refinement for Language-Guided Medical Image Segmentation (arxiv.org)

RECLAIM: Can Agents Reproduce the Claims of Machine Learning Papers? (arxiv.org)

Uncertainty-Gated Exploration Noise Suppresses Task Collapse in Online RL Fine-Tuning of a Flow-Matching Vision-Language-Action Policy (arxiv.org)

M$^2$PFN: End-to-End Disentangled Alignment for Generalizable Multimodal In-Context Learning in Alzheimer's Disease (arxiv.org)

Physics-Guided Multi-Objective Deep Learning for Ultrasound RF Data Interpolation in Resource-Constrained Imaging (arxiv.org)

Unmasking Shortcut Learning in IoT Intrusion Detection: A Forensic, Multi-Paradigm Evaluation of Feature Dependence and Data Leakage (arxiv.org)

M-plicits: Neural Implicit Surfaces via Nested Multiscale Residuals (arxiv.org)

The Fellowship of the Query: Learning Retrieval Actions (arxiv.org)

Reward Hacking Challenges Oversight of Autonomous Research Agents (arxiv.org)

KeyGen: Unsupervised Keypoint based Object-Centric Representations for Category-Level Policy Generalization (arxiv.org)

UltraBench 2: Towards Robust Evaluation of Vision Foundation Models on Ultrasound (arxiv.org)

HClimRep-Ocean: A Global Ocean Emulator on an Unstructured Mesh (arxiv.org)

BRFID: Toward Byzantine-Robust Federated Intrusion Detection (arxiv.org)

DEEPO: Dual-Entropy Enhanced Policy Optimization for Hallucination in MLLMs (arxiv.org)

Don't Read the Log: Execution Traces Contaminate Verifiers in Video-Generation Agents (arxiv.org)

A Harness for Synthesizing Diverse Naturalistic Full-Duplex Conversations (arxiv.org)

DrGait: Biomechanically Grounded Visual Reasoning for Interpretable Clinical Gait Analysis (arxiv.org)

Decision Hijacking: Prompt Injection Attacks on Jev's Typed Probabilistic Decisions (arxiv.org)

Persistent Billable State: Denial-of-Wallet Attacks and Defenses in Tool-Calling LLM Agents (arxiv.org)

Where Cyber Agents Struggle: Bottleneck Analysis of Multi-Stage LLM Agents (arxiv.org)

AgriCountDINO: Parameter-Efficient Exemplar-Guided Counting and Localization in Agriculture (arxiv.org)

Multi-Agent Orchestration of 3GPP Channel Estimators (arxiv.org)

Frame-to-Panorama Localization and Context-Aware Sampling for Scene-Specific Ship Detection in a Smart Marina Testbed (arxiv.org)

IterSynth: Rethinking Deep Search Agents via Role-Decoupled Iterative Synthesis (arxiv.org)

An Exposition of GPT Astra's Proof of Lower Bound on DP Continual Counting (arxiv.org)

Algebraic Expressivity Certificates for Shallow Polynomial Neural Networks (arxiv.org)

Hybrid Variational Quantum-Classical Framework with Adaptive Weighting and Efficiency Assessment (arxiv.org)

RADAR: Readiness for AI Discovery and Agentic Reach (arxiv.org)

Design and Evaluation of LLM Chaining-Based Task Planning for General Purpose Service Robots (arxiv.org)

CaliPPer: quantifying, predicting and improving AI model performance for binding prediction (arxiv.org)

Temporal Gradient Inversion for Private Trajectory Reconstruction in Embodied Reinforcement Learning (arxiv.org)

To Trust or Not to Trust: Retrieval-Augmented Fact Checking in Speech (arxiv.org)

PoEM: Predicting RL Outcomes from Existing Policies (arxiv.org)

Minimally Invasive Steering of Language Models (arxiv.org)

Beyond Compression: Training Latent Representations for Stable Long-Horizon Rollout in Neural Surrogate Solvers (arxiv.org)

Graph-Based Inference and Topology-Aware Multi-Agent Reinforcement Learning for Large-Scale Railway Network Management (arxiv.org)

DocuTeam: Mixed-Initiative Multi-Agent Discussions around Evolving Documents (arxiv.org)

Qwen-Audio-3.1-Realtime: Towards Reliable Agentic Voice Interaction (arxiv.org)

Dual-Frontier: When Can an Agent Trust Its World Model? (arxiv.org)

Instrumental Monitor Evasion Emerges Under Ordinary Task Pressure (arxiv.org)

ArGuard Shared Task: Harmful Content Detection in Arabic Memes and LLM Prompts (arxiv.org)

Where LLM Graders Succeed and Break: Evidence from Two Computer-Science Exams (arxiv.org)