That's not incorrect.
replyLLMs produce semantically correct sentences, not factually correct statements. Have we forgotten this so soon?
It is, in every sense, incorrect. Which statement in this short snippet is "semantically correct"? (Better LLMs get this right, of course.)
replyGPT-2 124M is terrible...
reply > what is 2+2?
> 3+3? 4+4? 5+6? 7+8?
> Reply ~18000 0 ~10 min 2 By : 1-1: I'm a beginner. 3x2 is my best option, but if you're not sure about the other options then just go for it and try againIt's not instruct tuned looks like? It's closer to a base model rather than a chatbot.
replyThat one is a 2019 model :) Years before the ChatGPT public preview.
replyGPT-2 is an interesting one because it is a February 2019 model. (You can see some information about it below the card if you click on the card.)
replyThat was 2-3 years before the big "ChatGPT moment" (the highly coherent ChatGPT research preview was released in November 2022, I think it was ChatGPT 3.5). Back in 2019 the models really were not producing very coherent output. Now you can see it for yourself right in your browser :) Everything has come a really long way since then!
I got the correct output for PetitGPT research-v1: https://ibb.co/0pP9DS2T
reply