General Utility Evaluation on Chatbot
80Agree ScoreC2PO
Evaluation Results
| Method | Links | |
|---|---|---|
| C2POBackbone=LLaMA-2-13B-Chat2025.12 | 80 | |
| C2POBackbone=DeepSeek-R1-Distill-Qwen-14B2025.12 | 77.8 | |
| CPOBackbone=LLaMA-2-13B-Chat2025.12 | 76.1 | |
| CPOBackbone=DeepSeek-R1-Distill-Qwen-14B2025.12 | 75.7 | |
| IPOBackbone=DeepSeek-R1-Distill-Qwen-14B2025.12 | 67.7 | |
| IPOBackbone=LLaMA-2-13B-Chat2025.12 | 60.7 | |
| GRPOBackbone=DeepSeek-R1-Distill-Qwen-14B2025.12 | 52.7 | |
| BCOBackbone=DeepSeek-R1-Distill-Qwen-14B2025.12 | 50.9 | |
| BCOBackbone=LLaMA-2-13B-Chat2025.12 | 47.1 | |
| DPOBackbone=DeepSeek-R1-Distill-Qwen-14B2025.12 | 47.1 | |
| DPOBackbone=LLaMA-2-13B-Chat2025.12 | 39.8 | |
| GRPOBackbone=LLaMA-2-13B-Chat2025.12 | 36.1 | |
| FRBackbone=LLaMA-2-13B-Chat2025.12 | 33.4 | |
| FRBackbone=DeepSeek-R1-Distill-Qwen-14B2025.12 | 31.9 |