List your company

Mathematical Reasoning on MATH 500 (Selection Accuracy)

63.87Selection AccuracyContrastive Causality (CC)
58.04659.55861.0762.582Jan 20, 2026

Evaluation results

MethodLinks
Contrastive Causality (CC)Perturbation=none2026.0163.87
Self-Certainty (SC)Perturbation=none2026.0163.8
Contrastive Causality (CC)Perturbation=shuffling2026.0162.13
Self-Certainty (SC)Perturbation=shuffling2026.0161.93
Self-Certainty (SC)Perturbation=small-eval2026.0159
Contrastive Causality (CC)Perturbation=small-eval2026.0158.27