Hacker News new | past | comments | ask | show | jobs | submit login

that resonates - less facts and more reasoning training data. The most low hanging in terms of non synthetic data probably being mathematical proofs. With prolog and the like many alternate reasoning paths could be generated. It's hard to say if these many-path would help in llm training without access to the gigantic machines (it's so unfair) to try it on.



Consider applying for YC's Fall 2025 batch! Applications are open till Aug 4

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

Search: