Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

What am I missing? How is any of his visualizations GPT-3 specific and not, say, a deep learning LSTM from years ago?


AFAIK, the only thing new about GPT-3 is its massive size, the architecture is completely conventional, so the same as those you've seen from a few years ago.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: