List your company

Human Preference Evaluation on PhysicsIQ 1.0 (test)

54.9Physics Plausibility Win RateMAGI-1
44.70847.3545052.646Jan 15, 2026

Evaluation results

MethodLinks
MAGI-1Sampling Strategy=WMReward(∇+BoN)2026.0154.958.752.957.155.764.954.560
vLDMSampling Strategy=WMReward(∇+BoN)2026.0153.154.854.758.851.353.25355.7
vLDMSampling Strategy=vanilla2026.0146.945.245.341.248.746.84744.3
MAGI-1Sampling Strategy=vanilla2026.0145.141.347.142.944.335.145.540