Machine Learning

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...

Accelerating science with AI and simulations

MIT Professor Rafael Gómez-Bombarelli discusses the transformative potential of AI in scientific research, emphasizing its role in materi...

Improving AI models’ ability to explain their predictions

[2603.26108] Accurate Precipitation Forecast by Efficiently Learning from Massive Atmospheric Variables and Unbalanced Distribution

Abstract page for arXiv paper 2603.26108: Accurate Precipitation Forecast by Efficiently Learning from Massive Atmospheric Variables and ...

[2603.26105] Are LLM-Enhanced Graph Neural Networks Robust against Poisoning Attacks?

Abstract page for arXiv paper 2603.26105: Are LLM-Enhanced Graph Neural Networks Robust against Poisoning Attacks?

[2603.26097] Dynamic Tokenization via Reinforcement Patching: End-to-end Training and Zero-shot Transfer

Abstract page for arXiv paper 2603.26097: Dynamic Tokenization via Reinforcement Patching: End-to-end Training and Zero-shot Transfer

[2603.26096] AcTTA: Rethinking Test-Time Adaptation via Dynamic Activation

Abstract page for arXiv paper 2603.26096: AcTTA: Rethinking Test-Time Adaptation via Dynamic Activation

[2603.26089] Selective Deficits in LLM Mental Self-Modeling in a Behavior-Based Test of Theory of Mind

Abstract page for arXiv paper 2603.26089: Selective Deficits in LLM Mental Self-Modeling in a Behavior-Based Test of Theory of Mind

[2603.26045] H-Node Attack and Defense in Large Language Models

Abstract page for arXiv paper 2603.26045: H-Node Attack and Defense in Large Language Models

[2603.25976] Second-Order, First-Class: A Composable Stack for Curvature-Aware Training

Abstract page for arXiv paper 2603.25976: Second-Order, First-Class: A Composable Stack for Curvature-Aware Training

[2603.25956] Adversarial-Robust Multivariate Time-Series Anomaly Detection via Joint Information Retention

Abstract page for arXiv paper 2603.25956: Adversarial-Robust Multivariate Time-Series Anomaly Detection via Joint Information Retention

[2603.25923] Preventing Data Leakage in EEG-Based Survival Prediction: A Two-Stage Embedding and Transformer Framework

Abstract page for arXiv paper 2603.25923: Preventing Data Leakage in EEG-Based Survival Prediction: A Two-Stage Embedding and Transformer...

[2603.25894] Data-Driven Plasticity Modeling via Acoustic Profiling

Abstract page for arXiv paper 2603.25894: Data-Driven Plasticity Modeling via Acoustic Profiling

[2603.25901] Decoding Defensive Coverage Responsibilities in American Football Using Factorized Attention Based Transformer Models

Abstract page for arXiv paper 2603.25901: Decoding Defensive Coverage Responsibilities in American Football Using Factorized Attention Ba...

[2603.25872] DRiffusion: Draft-and-Refine Process Parallelizes Diffusion Models with Ease

Abstract page for arXiv paper 2603.25872: DRiffusion: Draft-and-Refine Process Parallelizes Diffusion Models with Ease

[2603.25861] Why Safety Probes Catch Liars But Miss Fanatics

Abstract page for arXiv paper 2603.25861: Why Safety Probes Catch Liars But Miss Fanatics

[2603.25857] In-Context Molecular Property Prediction with LLMs: A Blinding Study on Memorization and Knowledge Conflicts

Abstract page for arXiv paper 2603.25857: In-Context Molecular Property Prediction with LLMs: A Blinding Study on Memorization and Knowle...

[2603.25839] A Compression Perspective on Simplicity Bias

Abstract page for arXiv paper 2603.25839: A Compression Perspective on Simplicity Bias

[2603.25813] MAGNET: Autonomous Expert Model Generation via Decentralized Autoresearch and BitNet Training

Abstract page for arXiv paper 2603.25813: MAGNET: Autonomous Expert Model Generation via Decentralized Autoresearch and BitNet Training

[2603.25779] Pure and Physics-Guided Deep Learning Solutions for Spatio-Temporal Groundwater Level Prediction at Arbitrary Locations

Abstract page for arXiv paper 2603.25779: Pure and Physics-Guided Deep Learning Solutions for Spatio-Temporal Groundwater Level Predictio...

[R] Editing ICML Rebuttal

Hi guys, If I submit my ICML rebuttal now on OpenReview, can I edit it afterwards until the deadline. submitted by /u/isentropiccombustor...

[R] I trained a 3k parameter model on XOR sequences of length 20. It extrapolates perfectly to length 1,000,000. Here's why I think that's architecturally significant.

I've been working on an alternative to attention-based sequence modeling that I'm calling Geometric Flow Networks (GFN). The core idea: i...

[P] I built an autonomous ML agent that runs experiments on tabular data indefinitely - inspired by Karpathy's AutoResearch

Inspired by Andrej Karpathy's AutoResearch, I built a system where Claude Code acts as an autonomous ML researcher on tabular binary clas...