List your company
ResearchFormerly WizWand

Mathematical Reasoning on GSM8K FR (test)

64AccuracyKL-DRTO
57.7659.386162.62Mar 27, 2026
Updated 6mo ago

Evaluation results

MethodLinks
KL-DRTO2026.0364
RTO2026.0360.8
DPO2026.0360.8
χ2-DRTO2026.0360.4
PPO2026.0359.6
SFT2026.0358.4
GRPO2026.0358