Go-Explore: a New Approach for Hard-Exploration Problems

preview_player
Показать описание
This algorithm solves the hardest games in the Atari suite and makes it look so easy! This modern version of Dijkstra's shortest path algorithm is outperforming everything else by orders of magnitude, and all based on random exploration.

Abstract:
A grand challenge in reinforcement learning is intelligent exploration, especially when rewards are sparse or deceptive. Two Atari games serve as benchmarks for such hard-exploration domains: Montezuma's Revenge and Pitfall. On both games, current RL algorithms perform poorly, even those with intrinsic motivation, which is the dominant method to improve performance on hard-exploration domains. To address this shortfall, we introduce a new algorithm called Go-Explore. It exploits the following principles: (1) remember previously visited states, (2) first return to a promising state (without exploration), then explore from it, and (3) solve simulated environments through any available means (including by introducing determinism), then robustify via imitation learning. The combined effect of these principles is a dramatic performance improvement on hard-exploration problems. On Montezuma's Revenge, Go-Explore scores a mean of over 43k points, almost 4 times the previous state of the art. Go-Explore can also harness human-provided domain knowledge and, when augmented with it, scores a mean of over 650k points on Montezuma's Revenge. Its max performance of nearly 18 million surpasses the human world record, meeting even the strictest definition of "superhuman" performance. On Pitfall, Go-Explore with domain knowledge is the first algorithm to score above zero. Its mean score of almost 60k points exceeds expert human performance. Because Go-Explore produces high-performing demonstrations automatically and cheaply, it also outperforms imitation learning work where humans provide solution demonstrations. Go-Explore opens up many new research directions into improving it and weaving its insights into current RL algorithms. It may also enable progress on previously unsolvable hard-exploration problems in many domains, especially those that harness a simulator during training (e.g. robotics).

Authors: Adrien Ecoffet, Joost Huizinga, Joel Lehman, Kenneth O. Stanley, Jeff Clune
Рекомендации по теме
Комментарии
Автор

Speed runners would widely approve of this method: replaying from saved states is about the only sane way to learn complex tasks.
Maybe you can view go explore as dynamic curriculums?

CristianGarcia
Автор

I don't think your explanation of Dijkstra's algorithm is accurate. As far as I know in an unweighted case it degenerates to Breadth-First Search, while your explanation looked more like Depth-First Search.
Otherwise a great video as always. Just found your channel and have been enjoying your videos immensly

hae_
Автор

Do you think the dimension reduction could also be learned? (14:30)
Without this I don't think it is that interesting for other problems. It would require a humanly crafted dimension reduction with domain knowledge.

risajef