List your company

Character Reasoning on Letter CD (4,5) (OOD)

14.4AccuracyR1 Distill -> GRPO
5.1447.5479.9512.353Dec 3, 2025

Evaluation results

MethodLinks
R1 Distill -> GRPOtraining=SFT + GRPO2025.1214.4
BOLT -> GRPOtraining=SFT + GRPO2025.1213.1
SkillFactory -> GRPOtraining=SFT + GRPO2025.1212.8
RL-Onlytraining=RL2025.1212.5
Qwen2.5 1.5B Instructtraining=None2025.1210.4
STaR -> GRPOtraining=SFT + GRPO2025.129.2
R1 Distilltraining=SFT2025.128.8
SkillFactorytraining=SFT2025.128.7
STaRtraining=SFT2025.127.3
BOLTtraining=SFT2025.125.5