[08] He He - Sequential Decisions and Predictions in NLP
The Thesis Review
The Relationship Between the Oracle and the Trust Region
In a game setting, you want to estimate how confident you are at the current position. And they also want to estimate whether the opponent or the other teams is going to answer. So that's why we started to think about this opponent modeling in this game setting. It makes sense in this competitive game setting where you're trying to compete with another one. But I think it's useful in collaborative settings as well. If you have all of their information, like you have a central optimizer that has access to all the agents state, you could do this pretty easily. When you only have partial view of the world, you don't know what the other agent is saying or planning to do.
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.
