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

I've tested the new model 4 here [1] to summarize research papers. It is still not enough - about 1500 - 3000 words can be fed in, depending on how many tokens are expected for the answer.

[1]: https://kartographie.geo.tu-dresden.de/ad/2022-12-22_OpenAI_...



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

Search: