CodeReclaimers

2 ms·

1. 1.▲
   by CodeReclaimers 4mo ago

   Agreed. The wrinkle I thought was worth writing up is: there's no learned reward model here and no training at all. The "reward" is wall-clock executiion time and the model is frozen; the search is happening at inference ti
2. 2.▲
   My LLM optimization loop reward-hacked its own benchmark (and other lessons) [pdf](github.com)

   1 points by CodeReclaimers 4mo ago|2 comments
3. 3.▲
   Show HN: Symbolic regression as an MCP tool (SINDy and PySR, free, no install)(occam.fit)

   5 points by CodeReclaimers 6mo ago|1 comments
4. 4.▲
   by CodeReclaimers 6mo ago

   Looking forward to the day when "Yesify is down, resulting in half the internet not working" is a real headline. :)