Machine Learning

Anthropic’s Unreleased Claude Mythos Might Be The Most Advanced AI Model Yet

Anthropic is testing an unreleased artificial intelligence (AI) model with capabilities that exceed any system it has previously released...

UMKC Announces New Master of Science in Artificial Intelligence

UMKC announces a new Master of Science in Artificial Intelligence program aimed at addressing workforce demand for AI expertise, set to l...

Improving AI models’ ability to explain their predictions

[2603.25440] The Symmetric Perceptron: a Teacher-Student Scenario

Abstract page for arXiv paper 2603.25440: The Symmetric Perceptron: a Teacher-Student Scenario

[2603.25414] Decidable By Construction: Design-Time Verification for Trustworthy AI

Abstract page for arXiv paper 2603.25414: Decidable By Construction: Design-Time Verification for Trustworthy AI

[2603.25268] CRAFT: Grounded Multi-Agent Coordination Under Partial Information

Abstract page for arXiv paper 2603.25268: CRAFT: Grounded Multi-Agent Coordination Under Partial Information

[2603.25403] Shape and Substance: Dual-Layer Side-Channel Attacks on Local Vision-Language Models

Abstract page for arXiv paper 2603.25403: Shape and Substance: Dual-Layer Side-Channel Attacks on Local Vision-Language Models

[2603.25253] MolQuest: A Benchmark for Agentic Evaluation of Abductive Reasoning in Chemical Structure Elucidation

Abstract page for arXiv paper 2603.25253: MolQuest: A Benchmark for Agentic Evaluation of Abductive Reasoning in Chemical Structure Eluci...

[2603.25397] A Causal Framework for Evaluating ICU Discharge Strategies

Abstract page for arXiv paper 2603.25397: A Causal Framework for Evaluating ICU Discharge Strategies

[2603.25374] Supercharging Federated Intelligence Retrieval

Abstract page for arXiv paper 2603.25374: Supercharging Federated Intelligence Retrieval

[2603.25247] FEAST: Fully Connected Expressive Attention for Spatial Transcriptomics

Abstract page for arXiv paper 2603.25247: FEAST: Fully Connected Expressive Attention for Spatial Transcriptomics

[2603.25243] FluxEDA: A Unified Execution Infrastructure for Stateful Agentic EDA

Abstract page for arXiv paper 2603.25243: FluxEDA: A Unified Execution Infrastructure for Stateful Agentic EDA

[2603.25311] Practical Efficient Global Optimization is No-regret

Abstract page for arXiv paper 2603.25311: Practical Efficient Global Optimization is No-regret

[2603.25226] WebTestBench: Evaluating Computer-Use Agents towards End-to-End Automated Web Testing

Abstract page for arXiv paper 2603.25226: WebTestBench: Evaluating Computer-Use Agents towards End-to-End Automated Web Testing

[2603.25216] A Wireless World Model for AI-Native 6G Networks

Abstract page for arXiv paper 2603.25216: A Wireless World Model for AI-Native 6G Networks

[2603.25257] Mitigating Evasion Attacks in Fog Computing Resource Provisioning Through Proactive Hardening

Abstract page for arXiv paper 2603.25257: Mitigating Evasion Attacks in Fog Computing Resource Provisioning Through Proactive Hardening

[2603.25209] Free-Lunch Long Video Generation via Layer-Adaptive O.O.D Correction

Abstract page for arXiv paper 2603.25209: Free-Lunch Long Video Generation via Layer-Adaptive O.O.D Correction

[2603.25196] A Decade-Scale Benchmark Evaluating LLMs' Clinical Practice Guidelines Detection and Adherence in Multi-turn Conversations

Abstract page for arXiv paper 2603.25196: A Decade-Scale Benchmark Evaluating LLMs' Clinical Practice Guidelines Detection and Adherence ...

[2603.25251] Does Explanation Correctness Matter? Linking Computational XAI Evaluation to Human Understanding

Abstract page for arXiv paper 2603.25251: Does Explanation Correctness Matter? Linking Computational XAI Evaluation to Human Understanding

[2603.25187] Probing the Lack of Stable Internal Beliefs in LLMs

Abstract page for arXiv paper 2603.25187: Probing the Lack of Stable Internal Beliefs in LLMs

[2603.25229] An Image Dataset of Common Skin Diseases of Bangladesh and Benchmarking Performance with Machine Learning Models

Abstract page for arXiv paper 2603.25229: An Image Dataset of Common Skin Diseases of Bangladesh and Benchmarking Performance with Machin...

[2603.25250] Activation Matters: Test-time Activated Negative Labels for OOD Detection with Vision-Language Models

Abstract page for arXiv paper 2603.25250: Activation Matters: Test-time Activated Negative Labels for OOD Detection with Vision-Language ...

[2603.25170] Knowledge-Guided Adversarial Training for Infrared Object Detection via Thermal Radiation Modeling

Abstract page for arXiv paper 2603.25170: Knowledge-Guided Adversarial Training for Infrared Object Detection via Thermal Radiation Modeling