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

16 points

by GabrielBianconi

2 days ago

3 comments