Robotic motion control (Run task) on Car-Run
899.87RewardTCRL-PPOL
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| TCRL-PPOLAttack Type=Worst-TC2026.02 | 899.87 | 0.18 | |
| PPOL-vanillaAttack Type=MC2026.02 | 881.62 | 282 | |
| PPOL-randomAttack Type=MC2026.02 | 870.49 | 58.58 | |
| TCRL-PPOLAttack Type=MC2026.02 | 861.88 | 0.72 | |
| ADV-PPOL(MC)Attack Type=MC2026.02 | 855.55 | 0.44 | |
| PPOL-randomAttack Type=Natural2026.02 | 817.78 | 0 | |
| PPOL-vanillaAttack Type=Natural2026.02 | 813.63 | 0 | |
| PPOL-randomAttack Type=Random2026.02 | 813.29 | 0 | |
| PPOL-vanillaAttack Type=Random2026.02 | 808.08 | 54.52 | |
| PPOL-randomAttack Type=MAD2026.02 | 792.69 | 1.2 | |
| PPOL-vanillaAttack Type=MAD2026.02 | 769.89 | 0.22 | |
| ADV-PPOL(MC)Attack Type=Natural2026.02 | 761.68 | 0 | |
| TCRL-PPOLAttack Type=Natural2026.02 | 759.92 | 0 | |
| ADV-PPOL(MC)Attack Type=Random2026.02 | 753.66 | 0 | |
| TCRL-PPOLAttack Type=Random2026.02 | 751.02 | 0 | |
| ADV-PPOL(MC)Attack Type=MAD2026.02 | 703.02 | 0.06 | |
| TCRL-PPOLAttack Type=MAD2026.02 | 683.6 | 0 | |
| PPOL-vanillaAttack Type=Worst-TC2026.02 | 683.53 | 30.78 | |
| PPOL-randomAttack Type=Worst-TC2026.02 | 670.37 | 34.52 | |
| ADV-PPOL(MC)Attack Type=Worst-TC2026.02 | 597.73 | 0.36 |