Meet the new biologists treating LLMs like aliens
MIT Technology Review Narrated
00:00
Designing more interpretable models
OpenAI considers training trade-offs to build easier-to-interpret models at reduced efficiency.
Play episode from 18:40
Transcript
