Weak-to-strong generalization (GPT-4 Turbo was Pareto-SOTA)

(forourposterity.com)

2 points | by ronfriedhaber 6 hours ago

No comments yet.