EQ-bench/EQ-Bench

default judge model setting for the leaderboard

Closed this issue · 1 comments

may I ask what the default judge model is?

For the creative writing leaderboard, it's claude-3-opus.

I will probably at some point make it an aggregate of multiple judges, since they all have a small amount of self-bias.

more info here: https://eqbench.com/about.html