List your company

Human Evaluation on NeurIPS ICML ICLR Proposals 2025 (test)

31WinsStepwise CoT
19.5622.5325.528.47Mar 28, 2026

Evaluation results

MethodLinks
Stepwise CoTComparison=Stepwise CoT vs. Prompting, Dimension=Overall2026.033132654.226.7
Stepwise CoTComparison=Stepwise CoT vs. Prompting, Dimension=Excitement2026.0330102058.326.7
Stepwise CoTComparison=Stepwise CoT vs. Prompting, Dimension=Soundness2026.032972454.231.7
Stepwise CoTComparison=Stepwise CoT vs. Human, Dimension=Overall2026.032510255011.7
Stepwise CoTComparison=Stepwise CoT vs. Human, Dimension=Soundness2026.0321112844.211.7
Stepwise CoTComparison=Stepwise CoT vs. Human, Dimension=Excitement2026.0320182248.36.7