Physical Reasoning on PHYRE-2B within-template (test)
77.6AUCCESSDynamics-Aware DQN
Evaluation results
| Method | Links | |
|---|---|---|
| Dynamics-Aware DQNLearning Mode=Self-supervised2021.02 | 77.6 | |
| Dynamics-Aware DQNLearning Mode=Hand-crafted2021.02 | 74 | |
| DQN2021.02 | 67.8 | |
| RAND2021.02 | 3.6 | |
| MEM2021.02 | 3.2 |