Bigger, Better, Faster: Human-level Atari with human-level efficiency

We introduce a value-based RL agent, which we call BBF, that achieves super-human performance in the Atari 100K benchmark. BBF relies on scaling the neural networks used for value estimation, as well as a number of other design choices that enable this scaling in a sample-efficient manner. We conduct extensive analyses of these design choices and provide insights for future work. We end with a discussion about updating the goalposts for sample-efficient RL research on the ALE. We make our code and data publicly available at https://github.com/google-research/google-research/tree/master/bigger_better_faster.

PDF Abstract

Datasets


Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Atari Games 100k Atari 100k BBF Mean Human-Normalized Score 2.245 # 1
Medium Human-Normalized Score 0.917 # 2

Methods


No methods listed for this paper. Add relevant methods here