The Inside View cover image

Collin Burns On Discovering Latent Knowledge In Language Models Without Supervision

The Inside View

CHAPTER

Is There a Difference Between Constraints and the Truth?

I think there are observational differences between like the truth and say what a human would say. If you put like enough constraints then maybe after some much musician we end up with truth. Now I'm imagining a more flexible model so instead of just doing something like fitting a probe on top of the activations instead I want to say something like let's search for something like a prompt or a prefix to the model so that it's output satisfies some properties. This is more flexible that means it's easier to find properties or find outputs or solutions that satisfy properties that we care about but my claim is first of all it is sufficient to get the model or like this the sort of model class this way of

00:00
Transcript
Play full episode

Remember Everything You Learn from Podcasts

Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.
App store bannerPlay store banner