galsapir

Recent Submissions

1. 1. ▲ Giving a domain a hill to climb: benchmarking as data activation (sparsethought.com) 12 points · 8 days ago · 7 comments
2. 2. ▲ A bitter lesson for medicine, or a benchmark problem? (sparsethought.com) 2 points · 27 days ago · 0 comments
3. 3. ▲ Can LLMs Beat Classical Hyperparameter Optimization Algorithms? (arxiv.org) 120 points · 1 month ago · 20 comments
4. 4. ▲ Gemma 4 E4B as a primary local LLM (replaced Qwen) (digg.com) 2 points · 1 month ago · 0 comments
5. 5. ▲ PEEK: Give Your Agent an Orientation Cache (MIT CSAIL, Khattab group) (zhuohangu.github.io) 3 points · 1 month ago · 0 comments
6. 6. ▲ Hyperagents (Meta Research) (arxiv.org) 2 points · 1 month ago · 0 comments
7. 7. ▲ The Unreasonable Effectiveness of HTML (claude.com) 3 points · 1 month ago · 2 comments
8. 8. ▲ The Comparator in Clinical AI (sparsethought.com) 2 points · 2 months ago · 1 comment