Even (very) noisy LLM evaluators are useful for improving AI agents
(tensorzero.com)
15 points
by GabrielBianconi
2 days ago |
1 comment
()
