List your company

Mathematics Reasoning on GSM8K (EM)

98.6Exact Match (EM)CoT2-Meta
94.02495.21296.497.588Mar 30, 2026

Evaluation results

MethodLinks
CoT2-MetaStrategy=Ours (CoT2-Meta), Inference Budget=C=162026.0398.6
ReST-MCTS*Strategy=ReST-MCTS*, Inference Budget=C=162026.0397.5
Vanilla ToTStrategy=Vanilla ToT, Inference Budget=C=162026.0396.8
Best-of-16Strategy=Best-of-16, Inference Budget=C=162026.0396.1
Greedy CoTStrategy=Greedy CoT, Inference Budget=C=162026.0394.2