List your company

Complaint Scene Reasoning on ComplaintScene-Text

83.26Avg Scoresft
56.386463.363270.3477.3168May 10, 2026

Evaluation results

MethodLinks
sftBackbone=Qwen2.5-3B2026.0583.2680.5576.9292.31
preBackbone=Qwen2.5-3B2026.0581.4378.6974.8190.79
sftBackbone=Qwen2.5-1.5B2026.0579.1475.8872.3689.18
preBackbone=Qwen2.5-1.5B2026.0576.5873.2469.1187.39
llmBackbone=Qwen2.5-3B2026.0574.6271.9566.8485.07
sftBackbone=Qwen2.5-0.5B2026.0571.4868.9565.7479.75
llmBackbone=Qwen2.5-1.5B2026.0569.8366.4161.2781.81
preBackbone=Qwen2.5-0.5B2026.0565.3161.0258.6476.27
llmBackbone=Qwen2.5-0.5B2026.0557.4254.1849.8768.21