List your company

Policy Referral on ClevrPolicy

6.26ClevrPolicy-T ScoreTriMPI
3.43124.16564.95.6344Oct 10, 2025

Evaluation results

MethodLinks
TriMPIStrategy=PoRo-DAPO2025.106.268.35
CoT SFT + DAPORL Algorithm=DAPO2025.105.76.74
TriMPIStrategy=PoRo-GRPO2025.105.68.72
CoT SFT + GRPORL Algorithm=GRPO2025.105.046.74
CoT SFT2025.103.543.44