List your company

Role-Playing Ability on Role-playing

4.99CUS ScorePPR-GDE
3.13883.61944.14.5806May 18, 2026

Evaluation results

MethodLinks
PPR-GDEBackbone=Qwen2.5-3B-Instruct2026.054.992.64.934.17
PPR-GDEBackbone=Qwen3-0.6B2026.054.972.254.944.05
PPR-GDEBackbone=Qwen2.5-1.5B-Instruct2026.054.942.824.864.21
GRPOBackbone=Qwen2.5-3B-Instruct2026.054.882.514.794.06
PPOBackbone=Qwen3-0.6B2026.054.832.624.794.08
GRPOBackbone=Qwen2.5-1.5B-Instruct2026.054.812.444.663.97
GRPOBackbone=Qwen3-0.6B2026.054.82.994.664.15
PPOBackbone=Qwen2.5-1.5B-Instruct2026.054.772.164.633.85
PPOBackbone=Qwen2.5-3B-Instruct2026.054.712.744.724.06
w/o RLBackbone=Qwen2.5-3B-Instruct2026.054.043.114.23.78
w/o RLBackbone=Qwen3-0.6B2026.053.381.793.352.84
w/o RLBackbone=Qwen2.5-1.5B-Instruct2026.053.2133.493.23