List your company

General Question Answering on UltraFeedback

0.1695RewardGRPO+START
0.1656520.1666510.167650.168649Mar 21, 2026

Evaluation results

MethodLinks
GRPO+STARTConfiguration=GRPO+START2026.030.169561.25
GRPOConfiguration=GRPO2026.030.1658—