Reinforcement Learning on Atari 2600 Kangaroo
11,200ScoreOpenAI ES
Evaluation results
| Method | Links | ||
|---|---|---|---|
| OpenAI ESinput type=raw pixel, # of neurons=~6502018.06 | 11,200 | — | |
| GA (1B)input type=raw pixel, # of neurons=~6502018.06 | 3,790 | — | |
| RainbowPlasticity=full, Activation=rational2021.02 | 2,157 | 6,000 | |
| RainbowPlasticity=regularised, Activation=joint-rational2021.02 | 2,139 | 4,800 | |
| IDVQ+DRSC+XNESinput type=raw pixel, max run length=200 interactions, frameskip=5, # of neurons=182018.06 | 1,200 | — | |
| HyperNeatinput type=raw pixel, # of neurons=~30342018.06 | 800 | — | |
| RainbowPlasticity=rigid, Activation=Leaky ReLU2021.02 | 40 | 6,300 |