List your company

Multi-turn Dialogue on NPC-Chat (test)

3.84FluencyAT-GRPO
2.9563.18553.4153.6445Feb 9, 2026

Evaluation results

MethodLinks
AT-GRPOEval. Type=LLM-based, Base Model=Qwen2.5-14B2026.023.843.823.413.82—3.72
GRPOEval. Type=LLM-based, Base Model=Qwen2.5-14B2026.023.833.823.373.76—3.7
SFTEval. Type=LLM-based, Base Model=Qwen2.5-14B2026.023.823.83.423.7—3.69
DeepSeekEval. Type=LLM-based, Base Model=Qwen2.5-14B2026.023.83.723.283.61—3.6
DeepSeekEval. Type=Human, Base Model=Qwen2.5-14B2026.023.13.042.842.563.523.01
GRPOEval. Type=Human, Base Model=Qwen2.5-14B2026.0233.012.92.724.63.25
AT-GRPOEval. Type=Human, Base Model=Qwen2.5-14B2026.0233.083.012.774.953.36
SFTEval. Type=Human, Base Model=Qwen2.5-14B2026.022.992.992.922.794.533.24