Even (very) noisy LLM evaluators are useful for improving AI agents
11 points
2 days ago
| 0 comments
| tensorzero.com
| HN
No one has commented on this post.