this post was submitted on 16 May 2023
1 points (66.7% liked)

Machine Learning

1796 readers
11 users here now

founded 4 years ago
MODERATORS
[email protected]
1
TinyStories: How Small Can Language Models Be and Still Speak Coherent English? (arxiv.org)
submitted 2 years ago by [email protected] to c/[email protected]
1 comments fedilink hide all child comments
you are viewing a single comment's thread
view the rest of the comments
[–] kraegar 1 points 2 years ago

NLP isn't an area I know much about, but I find it really interesting that they used GPT-4 as the scoring mechanism. I do have concerns about this after the MIT "100% OF QUESTIONS CORRECT" fiasco recently, but it will likely be as equally or less biased than a human marker.