Press Space to continue
Smarter Models Produce Better LLM Evaluations
Press Space to continue
Experimenting with automated evals, and whether more compute = better checks (LLM judges). Yes, kind of! Smarter models wrote better evals, and higher effort led to better evals for small models like Luna.
open on x · 1 viewViews and clicks count once per visitor per day. Clicks are opens of this post on X.