logoalt Hacker News

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

11 points • by GabrielBianconi • last Wednesday at 7:49 AM • 0 comments • view on HN

Comments