undefined | Better HN

0 pointsandyljones5y ago0 comments

It's worth chasing that with gwern's critique of Marcus' critque: https://www.gwern.net/GPT-3#marcus-2020

(the critique is: GPT-3 can in fact do all the things Marcus said it couldn't)

0 comments

4 comments · 1 top-level

sgt1015y ago· 3 in thread

I can't play with GPT-3 but when I play with GPT-2 I can easily trick it with counting games. It does well with 0,1,2,3,.... but things like 0,1,3,6,10, get poor responses. Is GPT-3 good at that?

sktguha5y ago

yes. I have tried it with gpt-3 Q: what comes next in the series: 0,3,6, A: 9

Q: what comes next in the series: 0,3,6,9, A: 12

ypcx5y ago

I think the question of "how much can it reason" can be made more specific as "how far away can it reason from the learned examples".

Increases in reasoning power should allow for much smaller usable models.

1 more reply

sgt1015y ago

I think that this is still reproduction. Try things like 1,A,3,C,5,E,7 or a,1,aa,2,aaa,3,aaaa

j / k navigate · click thread line to collapse