- cross-posted to:
- [email protected]
- cross-posted to:
- [email protected]
You must log in or # to comment.
It’s interesting how words with similar semantic content are close together in the abstract vector space. Those groupings seem like they could reveal the idological content of the training data.
Lol I don’t have it on hand, but someone did a centroid analysis of another smaller but open GPT model, and found the “word” at the “center” of the model space was “a man’s penis”
I found a YouTube link in your post. Here are links to the same video on alternative frontends that protect your privacy:
