List your company

Role-Playing on RPGBench Dialogue Shift (Generalization)

-0.956Turn CompositionRFT
-0.99248-0.74624-0.5-0.25376Dec 19, 2025

Evaluation results

MethodLinks
RFTBackbone=Qwen3-4B2025.12-0.956-0.583
DPOBackbone=Meta-Llama3-8B2025.12-0.565-0.387
ICLBackbone=Meta-Llama3-8B2025.12-0.415-0.014
ICLBackbone=Qwen3-4B2025.12-0.362-0.018
DPOBackbone=Qwen3-4B2025.12-0.292-0.111
RFTBackbone=Meta-Llama3-8B2025.12-0.241-0.127
ThinkingRLBackbone=Meta-Llama3-8B2025.12-0.218-0.059
ThinkingRLBackbone=Qwen3-4B2025.12-0.175-0.007
ThinkingSFTBackbone=Qwen3-4B2025.12-0.1450.078
ThinkingSFTBackbone=Meta-Llama3-8B2025.12-0.1050.116
RLBackbone=Meta-Llama3-8B2025.12-0.1010.164
DABackbone=Qwen3-4B2025.12-0.0980.137
SFTBackbone=Meta-Llama3-8B2025.12-0.0810.132
DABackbone=Meta-Llama3-8B2025.12-0.080.146
CoRLBackbone=Qwen3-4B2025.12-0.0660.153
RLBackbone=Qwen3-4B2025.12-0.0570.149
SFTBackbone=Qwen3-4B2025.12-0.0490.143
CoRLBackbone=Meta-Llama3-8B2025.12-0.0440.181