Commonsense Question Answering on CosmosQA (test)
92.25EMMetaTuner-J
Evaluation Results
| Method | Links | |
|---|---|---|
| MetaTuner-JBackbone=Qwen2.5-7B, Method Category=Hybrid2025.09 | 92.25 | |
| MetaTuner-IBackbone=Qwen2.5-7B, Method Category=Hybrid2025.09 | 92.15 | |
| BetterTogetherBackbone=Qwen2.5-7B, Method Category=Hybrid2025.09 | 89.8 | |
| DPOBackbone=Qwen2.5-7B, Method Category=Fine-Tuning2025.09 | 87.9 | |
| MetaTuner-JBackbone=Qwen2.5-3B, Method Category=Hybrid2025.09 | 87.15 | |
| BetterTogetherBackbone=Qwen2.5-3B, Method Category=Hybrid2025.09 | 86.6 | |
| MetaTuner-IBackbone=Qwen2.5-3B, Method Category=Hybrid2025.09 | 86.55 | |
| KTOBackbone=Qwen2.5-7B, Method Category=Fine-Tuning2025.09 | 85 | |
| CFPOBackbone=Qwen2.5-7B, Method Category=Prompt Optimization2025.09 | 84.8 | |
| PPOBackbone=Qwen2.5-7B, Method Category=Fine-Tuning2025.09 | 84.1 | |
| SFTBackbone=Qwen2.5-7B, Method Category=Fine-Tuning2025.09 | 82.65 | |
| BPOBackbone=Qwen2.5-7B, Method Category=Prompt Optimization2025.09 | 82.05 | |
| RLPromptBackbone=Qwen2.5-7B, Method Category=Prompt Optimization2025.09 | 81.2 | |
| BPOBackbone=Qwen2.5-3B, Method Category=Prompt Optimization2025.09 | 72.9 | |
| OPROBackbone=Qwen2.5-3B, Method Category=Prompt Optimization2025.09 | 69.35 | |
| OPROBackbone=Qwen2.5-7B, Method Category=Prompt Optimization2025.09 | 69.1 | |
| RLPromptBackbone=Qwen2.5-3B, Method Category=Prompt Optimization2025.09 | 67.15 | |
| CFPOBackbone=Qwen2.5-3B, Method Category=Prompt Optimization2025.09 | 64.2 | |
| DPOBackbone=Qwen2.5-3B, Method Category=Fine-Tuning2025.09 | 61.2 | |
| KTOBackbone=Qwen2.5-3B, Method Category=Fine-Tuning2025.09 | 58.7 | |
| PPOBackbone=Qwen2.5-3B, Method Category=Fine-Tuning2025.09 | 52.25 | |
| SFTBackbone=Qwen2.5-3B, Method Category=Fine-Tuning2025.09 | 52.05 | |
| Qwen2.5Backbone=Qwen2.5-7B, Method Category=Vanilla2025.09 | 36.8 | |
| Qwen2.5Backbone=Qwen2.5-3B, Method Category=Vanilla2025.09 | 25 |