ML Newsnew | past | best | rss | submit about
Stories from April 30, 2026 (UTC)
Go back a day, month, or year. Go forward a day.
1. Fast GPU Linear Algebra via Compile Time Expression Fusion (arxiv.org)
12 points by matt_d 156 days ago | hide | past | pdf | discuss
2. LLMs Corrupt Your Documents When You Delegate (arxiv.org)
4 points by uxhacker 156 days ago | hide | past | pdf | discuss
3. Estimating Black-Box LLM Parameter Counts via Factual Capacity (arxiv.org)
3 points by RockstarSprain 156 days ago | hide | past | pdf | discuss
4. Beyond 80/20: High-Entropy Minority Tokens Drive Effective RL for LLM Reasoning (arxiv.org)
3 points by mdp2021 156 days ago | hide | past | pdf | discuss
5. MathDuels: Evaluating LLMs as Problem Posers and Solvers (arxiv.org)
2 points by matt_d 156 days ago | hide | past | pdf | 1 comment
6. Outcome Rewards Do Not Guarantee Verifiable or Causally Important Reasoning (arxiv.org)
2 points by krackers 156 days ago | hide | past | pdf | 1 comment
7. Performance Analysis of AI Query Approximation Using Lightweight Proxy Models (arxiv.org)
1 point by tanelpoder 156 days ago | hide | past | pdf | discuss
8. Kernel Contracts: A Spec. Language for Correctness Across Heterogeneous Silicon (arxiv.org)
1 point by matt_d 156 days ago | hide | past | pdf | discuss
9. Prism: Demystifying Retention and Interaction in Mid-Training (arxiv.org)
1 point by mdp2021 156 days ago | hide | past | pdf | discuss
About | RSS | RSS (all) | HN arXiv