List your company

Mathematical Reasoning on GSM8K en (Accuracy)

78.42Accuracy (GSM8K en)Baseline
58.587263.736168.88574.0339Apr 29, 2026

Evaluation results

MethodLinks
BaselineN=16, setting=English occurrence as language confusion, aggregation=Averaged across four models and four target languages2026.0478.42
TLPON=16, setting=English occurrence as language confusion, aggregation=Averaged across four models and four target languages2026.0477.86
DPON=16, setting=English occurrence as language confusion, aggregation=Averaged across four models and four target languages2026.0475.04
ORPON=16, setting=English occurrence as language confusion, aggregation=Averaged across four models and four target languages2026.0472.86
SFTN=16, setting=English occurrence as language confusion, aggregation=Averaged across four models and four target languages2026.0459.35