Atari Games On Atari Game
评估指标
Human World Record Breakthrough
评测结果
各个模型在此基准测试上的表现结果
模型名称 | Human World Record Breakthrough | Paper Title | Repository |
|---|---|---|---|
| Go-Explore | 17 | Go-Explore: a New Approach for Hard-Exploration Problems | |
| Agent57 | 18 | Agent57: Outperforming the Atari Human Benchmark | |
| GDI-H3 | 22 | Generalized Data Distribution Iteration | - |
| R2D2 | 15 | Recurrent Experience Replay in Distributed Reinforcement Learning | - |
| Rainbow | 4 | Rainbow: Combining Improvements in Deep Reinforcement Learning | |
| NGU | 8 | Never Give Up: Learning Directed Exploration Strategies | |
| Muesli | 5 | Muesli: Combining Improvements in Policy Optimization | |
| GDI-I3 | 17 | Generalized Data Distribution Iteration | - |
| Muzero | 19 | Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model |
0 of 9 row(s) selected.