Reinforcement Learning on Atari 2600 Seaquest
4,770Average ScoreRUDDER
Evaluation Results
| Method | Links | |
|---|---|---|
| RUDDERDelay (frames)=272, Delay-event=collect divers, Training game frames=200M, Number of random seeds=3, Trials per seed=10, Frame skip=42018.06 | 4,770 | |
| OPRIDEObservation=pixel-based, Human-in-the-loop=true2026.02 | 3,478.4 | |
| IDRLObservation=pixel-based, Human-in-the-loop=true2026.02 | 2,941.2 | |
| OPRLObservation=pixel-based, Human-in-the-loop=true2026.02 | 2,784.1 | |
| PTObservation=pixel-based, Human-in-the-loop=true2026.02 | 2,538.3 | |
| PT+PDSObservation=pixel-based, Human-in-the-loop=true2026.02 | 2,459.6 | |
| PPODelay (frames)=272, Delay-event=collect divers, Training game frames=200M, Number of random seeds=3, Trials per seed=10, Frame skip=42018.06 | 1,616 | |
| OpenAI ESinput type=raw pixel, # of neurons=~6502018.06 | 1,390 | |
| NSRA-ESinput type=Atari RAM, # of neurons=~6502018.06 | 960 | |
| GA (1B)input type=raw pixel, # of neurons=~6502018.06 | 798 | |
| HyperNeatinput type=raw pixel, # of neurons=~30342018.06 | 716 | |
| IDVQ+DRSC+XNESinput type=raw pixel, max run length=200 interactions, frameskip=5, # of neurons=182018.06 | 320 |