Action-component payoff optimization on L6 warmup discriminative cell
3.94Per-interaction PayoffReciprocity Gradient
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Reciprocity GradientTouter=400, Seeds=102026.05 | 3.94 | 87 | |
| DPGTouter=125, Seeds=102026.05 | 2.19 | 49 | |
| TD3Touter=125, Seeds=102026.05 | 2.1 | 47 | |
| DDPGTouter=125, Seeds=102026.05 | 2.09 | 46 |