List your company

Mathematical Reasoning on GSM8K (SC Accuracy & Efficiency Metrics)

90.9SC AccuracySelf-Consistency
71.86876.80981.7586.691Jan 29, 2026

Evaluation results

MethodLinks
Self-ConsistencyModel=QWQ32B, Sample size N=312026.0190.9———
PoLRModel=QWQ32B, Sample size N=312026.0190.9-0.648.65.8
Self-ConsistencyModel=QWQ32B, Sample size N=512026.0190.8———
PoLRModel=QWQ32B, Sample size N=512026.0190.8-0.347.611.2
Self-ConsistencyModel=QWQ32B, Sample size N=112026.0190.6———
PoLRModel=QWQ32B, Sample size N=112026.0190.6-1.354.22.4
Self-ConsistencyModel=DSQ7B, Sample size N=312026.0180———
PoLRModel=DSQ7B, Sample size N=312026.0180-0.427.24.3
Self-ConsistencyModel=DSQ7B, Sample size N=512026.0179.8———
PoLRModel=DSQ7B, Sample size N=512026.0179.80.226.55.9
Self-ConsistencyModel=DSQ7B, Sample size N=112026.0179.7———
PoLRModel=DSQ7B, Sample size N=112026.0179.7028.11.9
Self-ConsistencyModel=DSQ1.5B, Sample size N=512026.0173.2———
PoLRModel=DSQ1.5B, Sample size N=512026.0173.2040.111.3
Self-ConsistencyModel=DSQ1.5B, Sample size N=312026.0173.2———
PoLRModel=DSQ1.5B, Sample size N=312026.0173.2041.16
Self-ConsistencyModel=DSQ1.5B, Sample size N=112026.0172.6———
PoLRModel=DSQ1.5B, Sample size N=112026.0172.6043.72.3