The Gradient: Perspectives on AI cover image

Melanie Mitchell: Abstraction and Analogy in AI

The Gradient: Perspectives on AI

CHAPTER

How Do We Evaluate Language Models?

I suppose that it's really only been the past decade or so where we've seen there be enough data and compute in order to allow researchers to leverage. I suppose a big aspect of it is that now we can learn some empirical weight, I think, to some of these questions that maybe we couldn't before. To some extent, yes, although I'm not sure we have the right metrics for this empirical study. So when you look at language models and how do we evaluate them, how do we say that we're making progress? Well, there's a few different ways.

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