AXRP - the AI X-risk Research Podcast cover image

16 - Preparing for Debate AI with Geoffrey Irving

AXRP - the AI X-risk Research Podcast

CHAPTER

Teaching Language Models to Support Answers With Verified Quotes

The goal of this work is to make models more ffactual. We want models to be accurate. And just showing a human an answer to some random factual quest like how long did george washington live, is is stupid. What you should actually do as you you the model, or through some process, you have to get the human information that it that the human can trust. This approaches trying to make the model do that quotation process itself. So stead of it's a question antr system, you take a question, the model replies with, ah, it's sort of concise answer. And then a a segment from a i rom page on the internet astead ofa ver

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