List your company

Reward Performance on Webshop Agent High-data

45Test-Time RewardMC
26.2831.143640.86May 23, 2026

Evaluation results

MethodLinks
MCtraining_regime=BOOST2026.0545
MCtraining_regime=Bilevel Only2026.0543
ILQLtraining_regime=Bilevel Only2026.0541
MCtraining_regime=Vanilla2026.0538
ILQLtraining_regime=BOOST2026.0536
MCtraining_regime=Vanilla+Syn2026.0531
ILQLtraining_regime=Vanilla2026.0531
ILQLtraining_regime=Vanilla+Syn2026.0527