Text Summarization on Summarize
67.39Average ScoreCFA
Evaluation Results
| Method | Links | |
|---|---|---|
| CFAModel Backbone=Llama2-7B, Alignment Algorithm=PPO2026.01 | 67.39 | |
| CFAModel Backbone=Llama2-7B, Alignment Algorithm=DPO2026.01 | 67.3 | |
| PPOModel Backbone=Llama2-7B, Alignment Strategy=Base2026.01 | 65.88 | |
| DPOModel Backbone=Llama2-7B, Alignment Strategy=Base2026.01 | 65.68 | |
| SFTModel Backbone=Llama2-7B2026.01 | 63.91 | |
| SFTModel Backbone=Llama2-7B2026.01 | 63.91 | |
| CFAModel Backbone=Llama3.1-8B, Alignment Algorithm=PPO2026.01 | 57.59 | |
| CFAModel Backbone=Llama3.1-8B, Alignment Algorithm=DPO2026.01 | 57.44 | |
| PPOModel Backbone=Llama3.1-8B, Alignment Strategy=Base2026.01 | 56.93 | |
| DPOModel Backbone=Llama3.1-8B, Alignment Strategy=Base2026.01 | 56.7 | |
| SFTModel Backbone=Llama3.1-8B2026.01 | 56.18 | |
| SFTModel Backbone=Llama3.1-8B2026.01 | 56.18 | |
| CFAModel Backbone=Qwen2.5-7B, Alignment Algorithm=PPO2026.01 | 55.21 | |
| CFAModel Backbone=Qwen2.5-7B, Alignment Algorithm=DPO2026.01 | 55.01 | |
| PPOModel Backbone=Qwen2.5-7B, Alignment Strategy=Base2026.01 | 54.29 | |
| DPOModel Backbone=Qwen2.5-7B, Alignment Strategy=Base2026.01 | 53.85 | |
| SFTModel Backbone=Qwen2.5-7B2026.01 | 52.42 | |
| SFTModel Backbone=Qwen2.5-7B2026.01 | 52.42 |