Natasha Jaques 2
TalkRL: The Reinforcement Learning Podcast
Token Level Probabilities
RL normally is like their discount factor is one. So they apply the same reward to all of the tokens in the sequence. But once you're already having a long conversation, then the reward is higher. Because it matters just as much what you say at the end.
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.
