List your company

Reward Performance on Twenty Questions High-data

5.83Test-Time RewardILQL
1.88842.91173.9354.9583May 23, 2026

Evaluation results

MethodLinks
ILQLtraining_regime=BOOST2026.055.83
ILQLtraining_regime=Bilevel Only2026.055.38
MCtraining_regime=Bilevel Only2026.053.83
MCtraining_regime=BOOST2026.053.48
ILQLtraining_regime=Vanilla2026.052.6
MCtraining_regime=Vanilla2026.052.22
MCtraining_regime=Vanilla+Syn2026.052.19
ILQLtraining_regime=Vanilla+Syn2026.052.04