pathak22 / noreward-rl

[ICML 2017] TensorFlow code for Curiosity-driven Exploration for Deep Reinforcement Learning

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Does this work on Atari?

seann999 opened this issue · comments

The paper goes over Mario and Doom, but if any experiments were done on Atari, it'd nice to know.

I did not experiment on Atari as much because there is no notion of generalization across levels. All levels are same in Atari. Moreover most games in Atari are dense reward, so not sure how much is going to help. The hyper-parameters would need tuning though.