List your company

Language Adherence on GSM8K (cross)

99.72RPRSFT
97.998.372598.84599.3175Apr 29, 2026

Evaluation results

MethodLinks
SFTEvaluation Setting=English as a Neutral Category2026.0499.7299.99
TLPOEvaluation Setting=English as a Neutral Category, N=162026.0499.5899.99
DPOEvaluation Setting=English as a Neutral Category2026.0499.0299.83
ORPOEvaluation Setting=English as a Neutral Category2026.0498.6399.9
BaselineEvaluation Setting=English as a Neutral Category2026.0497.9799.92