List your company

Instruction Following on IFEval (Acc_prompt)

19.6Accuracy (Prompt)FiMi-RM
16.4817.2918.118.91May 19, 2025

Evaluation results

MethodLinks
FiMi-RMBackbone=Gemma2-9B, Optimization Algorithm=DPO2025.0519.6
FiMi-RMBackbone=Qwen2.5-7B, Optimization Algorithm=DPO2025.0519
ODINBackbone=Qwen2.5-7B, Optimization Algorithm=DPO2025.0518.6
Length PenaltyBackbone=Gemma2-9B, Optimization Algorithm=DPO2025.0518.2
Vanilla RMBackbone=Gemma2-9B, Optimization Algorithm=DPO2025.0518
ODINBackbone=Gemma2-9B, Optimization Algorithm=DPO2025.0517.8
Vanilla RMBackbone=Qwen2.5-7B, Optimization Algorithm=DPO2025.0517.4
Length PenaltyBackbone=Qwen2.5-7B, Optimization Algorithm=DPO2025.0516.6