Episode 19: Minqi Jiang, UCL, on environment and curriculum design for general RL agents
Generally Intelligent
00:00
Is There Any Way to Improve the Performance of Your Agent?
The number one thing that i think is like the most helpful, just generally, for doing a i stuff, is basely just print everything out. One example of this is doing the robus pular paper. I trained a bunch of car raicing agents on the cluster and then printed out the actual behavior. And because of seeng essential, effectively different representations than saw during training. That's another example where you don't really know until you actually look at it to later.
Transcript
Play full episode
