Model Editing on FEVER
98.23EfficacyUltraEdit*
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| UltraEdit*Backbone=Mistral-7B-v0.3, Edit Scenario=ultra-large-scale2025.05 | 98.23 | 96.97 | 83.43 | |
| UltraEdit*Backbone=Mistral-7B-v0.32025.05 | 98.23 | 96.97 | 83.43 | |
| UltraEditModel=Qwen2.5-7B-Instruct2025.05 | 97.97 | 93.91 | 68.86 | |
| UltraEdit*Backbone=GPT-J, Edit Scenario=ultra-large-scale2025.05 | 97.89 | 96.73 | 79.85 | |
| UltraEdit*Backbone=GPT-J2025.05 | 97.89 | 96.73 | 79.85 | |
| UltraEditBackbone=Mistral-7B-v0.3, Edit Scenario=standard2025.05 | 97.87 | 96.09 | 84.29 | |
| UltraEditBackbone=Mistral-7B-v0.32025.05 | 97.87 | 96.09 | 84.29 | |
| UltraEdit*Model=Qwen2.5-7B-Instruct2025.05 | 97.49 | 94.83 | 68.99 | |
| UltraEditBackbone=GPT-J, Edit Scenario=standard2025.05 | 97.45 | 96.37 | 79.72 | |
| UltraEditBackbone=GPT-J2025.05 | 97.45 | 96.37 | 79.72 | |
| UltraEdit*Backbone=LLaMA-3-8B-Instruct, Edit Scenario=ultra-large-scale2025.05 | 97.18 | 94.64 | 68.62 | |
| UltraEdit*Backbone=LLaMA-3-8B-Instruct2025.05 | 97.18 | 94.64 | 68.62 | |
| UltraEditBackbone=LLaMA-3-8B-Instruct, Edit Scenario=standard2025.05 | 95.39 | 91.93 | 67.14 | |
| UltraEditBackbone=LLaMA-3-8B-Instruct2025.05 | 95.39 | 91.93 | 67.14 | |
| MALMENBackbone=LLaMA-3-8B-Instruct2025.05 | 94.5 | 91.26 | 67.76 | |
| RLEditBackbone=LLaMA-3-8B-Instruct, Edit Scenario=standard2025.05 | 94.03 | 90.67 | 68.71 | |
| WISEBackbone=GPT-J, Edit Scenario=standard2025.05 | 93.95 | 92.86 | 57.03 | |
| RLEditBackbone=Mistral-7B-v0.3, Edit Scenario=standard2025.05 | 92.35 | 91.39 | 71.85 | |
| WISEBackbone=LLaMA-3-8B-Instruct, Edit Scenario=standard2025.05 | 86.38 | 86.36 | 72.08 | |
| WISEBackbone=Mistral-7B-v0.3, Edit Scenario=standard2025.05 | 81.95 | 76.94 | 40.37 | |
| MENDModel=Qwen2.5-7B-Instruct2025.05 | 76.43 | 77.66 | 40.39 | |
| MENDBackbone=GPT-J2025.05 | 52.8 | 51.44 | 45.42 | |
| MENDBackbone=LLaMA-3-8B-Instruct2025.05 | 36.19 | 36.19 | 24.31 | |
| AlphaEditModel=Qwen2.5-7B-Instruct2025.05 | 32.78 | 31.19 | 22.12 | |
| FTModel=Qwen2.5-7B-Instruct2025.05 | 26.09 | 24.62 | 21.36 | |
| FTBackbone=Mistral-7B-v0.3, Edit Scenario=standard2025.05 | 23.8 | 23.37 | 16.3 | |
| MALMENBackbone=Mistral-7B-v0.32025.05 | 18.42 | 17.43 | 12.09 | |
| RECTBackbone=GPT-J2025.05 | 18.12 | 18.08 | 12.27 | |
| FTBackbone=LLaMA-3-8B-Instruct, Edit Scenario=standard2025.05 | 16.21 | 13.08 | 5.01 | |
| RLEditBackbone=GPT-J, Edit Scenario=standard2025.05 | 14.26 | 13.74 | 14.11 | |
| FTBackbone=GPT-J, Edit Scenario=standard2025.05 | 14.02 | 13.98 | 9.26 | |
| AlphaEditBackbone=LLaMA-3-8B-Instruct, Edit Scenario=standard2025.05 | 6.39 | 6.14 | 2.72 | |
| MEMITBackbone=GPT-J2025.05 | 5.54 | 5.03 | 5.46 | |
| PRUNEBackbone=GPT-J2025.05 | 5.25 | 4.72 | 5.2 | |
| RECTModel=Qwen2.5-7B-Instruct2025.05 | 3.36 | 3.1 | 3.2 | |
| AlphaEditBackbone=GPT-J, Edit Scenario=standard2025.05 | 1.89 | 1.87 | 1.85 | |
| MALMENBackbone=GPT-J2025.05 | 1.33 | 1.25 | 2.92 | |
| MEMITModel=Qwen2.5-7B-Instruct2025.05 | 0.08 | 0.12 | 0.15 | |
| PRUNEModel=Qwen2.5-7B-Instruct2025.05 | 0.08 | 0.15 | 0.11 | |
| MALMENModel=Qwen2.5-7B-Instruct2025.05 | 0.06 | 0.06 | 0.07 | |
| MEMITBackbone=LLaMA-3-8B-Instruct2025.05 | 0.02 | 0.02 | 0.02 | |
| PRUNEBackbone=LLaMA-3-8B-Instruct2025.05 | 0.02 | 0.02 | 0 | |
| RECTBackbone=LLaMA-3-8B-Instruct2025.05 | 0.01 | 0 | 0 | |
| AlphaEditBackbone=Mistral-7B-v0.3, Edit Scenario=standard2025.05 | 0 | 0 | 0 | |
| RLEditModel=Qwen2.5-7B-Instruct2025.05 | 0 | 0 | 0 | |
| MENDBackbone=Mistral-7B-v0.32025.05 | 0 | 0 | 0 | |
| MEMITBackbone=Mistral-7B-v0.32025.05 | 0 | 0 | 0 | |
| RECTBackbone=Mistral-7B-v0.32025.05 | 0 | 0 | 0 | |
| PRUNEBackbone=Mistral-7B-v0.32025.05 | 0 | 0 | 0 |