Evaluating GPT-3 and GPT-4 on the Winograd Schema Challenge(d-kz.medium.com)

1 pointsby nicclقبل 3 سنوات2 comments

https://d-kz.medium.com/evaluating-gpt-3-and-gpt-4-on-the-winograd-schema-challenge-reasoning-test-e4de030d190d

1 comments

sharemywinقبل 3 سنوات

I'm guessing they trained it on most if not all of those datasets.

qualudeheartقبل 3 سنوات

Maybe. Unclear if that’s falsifiable.