Hacker News   << back
code
Even (very) noisy LLM evaluators are useful for improving AI agents (www.tensorzero.com)
10 points by GabrielBianconi 2 days ago | 0 comments