Danau5tin | Dark Hacker News

1.	Show HN: I RL-trained an agent that trains models with RL (for –$1.3k)(github.com) 59 points by Danau5tin 4 hours ago \| 23 comments
2.	1 points by Danau5tin 230 days ago \| discuss
3.	Scaling Coding-Agent RL to 32x H100s. 160% Improvement on Stanford's TBench(github.com) 2 points by Danau5tin 253 days ago \| 1 comment
4.	Show HN: Multi-Agent-Coder Is #12 on Stanford's TBench. Beats Claude Code(github.com) 5 points by Danau5tin 314 days ago \| 1 comment
5.	My weekend project accidentally beat Claude Code – #12 on Stanford's TBench(github.com) 2 points by Danau5tin 315 days ago \| 2 comments
6.	Show HN: Terminal-Bench-RL: Training long-horizon terminal agents with RL(github.com) 125 points by Danau5tin 350 days ago \| 12 comments