Even (very) noisy LLM evaluators are useful for improving AI agents

(tensorzero.com)

10 points | by GabrielBianconi 2 days ago ago

No comments yet.