Reward Performance on Car Negotiation High-data
41.42Test-Time RewardMC
Evaluation Results
| Method | Links | |
|---|---|---|
| MCtraining_regime=BOOST2026.05 | 41.42 | |
| MCtraining_regime=Bilevel Only2026.05 | 41.31 | |
| ILQLtraining_regime=Bilevel Only2026.05 | 37.55 | |
| ILQLtraining_regime=BOOST2026.05 | 37.36 | |
| ILQLtraining_regime=Vanilla2026.05 | 35.33 | |
| MCtraining_regime=Vanilla+Syn2026.05 | 35.09 | |
| ILQLtraining_regime=Vanilla+Syn2026.05 | 33 | |
| MCtraining_regime=Vanilla2026.05 | 32.27 |