Preference Classification on Instruction Following (IF) (test)
61.4AccuracyBTPO
Evaluation Results
| Method | Links | |
|---|---|---|
| BTPOBackbone=Qwen2.5-7B-Instruct2025.10 | 61.4 | |
| RM-R1Backbone=Qwen2.5-7B-Instruct2025.10 | 52.2 | |
| RM-R1-finetunedBackbone=Qwen2.5-7B-Instruct, Fine-tuned=true2025.10 | 52.2 | |
| GRPO (pair)Backbone=Qwen2.5-7B-Instruct2025.10 | 52.2 |