Trading Backtest on AAPL held-out one-hour (test)
1,925.01Average ReturnGRPO
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| GRPO2026.05 | 1,925.01 | 6,408.17 | 97.69 | 72.86 | -1,235.17 | |
| GSPO2026.05 | 1,835.91 | 6,059.83 | 51.82 | 70.82 | -1,428.34 | |
| Q-Learning2026.05 | 19.24 | 1,404.8 | 5.61 | 15.41 | -53,564.33 | |
| PPO2026.05 | -44.33 | 1,538.85 | 1.66 | 34.98 | -12,346.35 |