1 Comment

Have you tried running your prompts with the new o1 models? It gave me 1% for the nuclear question.

Expand full comment