Thread

1. Garlef · 1791090364 · focus · HN ↗

   I think they even more so need deterministic feedback:

   I tried an approach based on the following idea recently and it's amazing - Lint rules where the error messages contain an explanation on how to deal with the issue.

   <a href="https:&#x2F;&#x2F;habit-hooks.com&#x2F;" rel="nofollow">https:&#x2F;&#x2F;habit-hooks.com&#x2F;

   I&#x27;m using it to foster IOSP (integration operation segregation principle) for example.

   1. fxtentacle · 1791101965 · focus · HN ↗

      Deterministic feedback is precisely how frontier models are trained. It’s called RLVR. You let the agent run on a problem and then calculate a deterministic score of how well it did. Repeat 1000x times and you can “brute force” a good solution. (Which includes all thinking traces and you add it to your training data.) And then a Chinese model can copy your advance for 1000x less compute. Which is why US labs call this not learning, but a distillation “attack”. It’s an attack on the business model.

      1. balder1991 · 1791125605 · focus · HN ↗

         I suppose the same way that the normal programmers and artists would call LLMs an attack on licenses and copyright.

Unofficial Hacker News client; not affiliated with Y Combinator.