News for hackers. Stuff that matters.

A comment on GPT-4. Parent comment

Re:GPT-4 by pja on Tuesday March 14, 2023 @03:21PM (#35156999)
More accurately: a GPT derived DNN that’s been specifically trained (or fine-tuned, if you want to use OpenAI’s language) on a dataset of Othello games ends up with an internal model of an Othello board.

It looks like OpenAI have specifically added Othello game handling to chat.openai.org, so I guess they’ve done the same fine-tuning to ChatGPT? It would be interesting to know how good an untuned GPT3/4 was at Othello & whether OpenAI has fine-tuned it or not!

(Having just tried a few moves, it looks like ChatGPT is just as bad at Othello as it was at chess, so it’s interesting that it knows the initial board layout but can’t actually play any moves correctly: Every updated board it prints out is completely wrong.)

[ Parent | on HN ]
Re:GPT-4 by WoodenChair on Tuesday March 14, 2023 @05:14PM (#35158785)
> it’s interesting that it knows the initial board layout

Why is that interesting? The initial board layout would appear all the time in the training data.

[ Parent | on HN ]
Re:GPT-4 by brokensegue on Wednesday March 15, 2023 @02:25AM (#35164620)
the initial board state is not ever encoded in the representation they use. imagine deducing the initial state of a chess board from the sequence of moves.
[ Parent | on HN ]