Hacker News new | ask | show | jobs
by kromem 752 days ago
I have skepticism regarding the 'completeness' of SAE in comprehensive discovery of features:

https://www.lesswrong.com/posts/BduCMgmjJnCtc7jKc/research-r...
1 comments
jengels_ 752 days ago
Sure, but completeness is a much higher bar than being able to find at least some things we weren’t looking for. And I’m reasonably optimistic that we’re going to make SAEs much better in the future, I agree they’re definitely imperfect right now
link