Knowledge Editing on Counterfact uns
94.56Edit Success RateCoT2Edit
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| CoT2EditModel=Falcon3-10B2026.04 | 94.56 | 81.43 | 90.61 | |
| CoT2EditModel=Qwen3-14B2026.04 | 94.46 | 81.02 | 84.51 | |
| CoT2EditModel=LLaMA3-8B2026.04 | 93.64 | 79.79 | 93.05 | |
| LTEModel=Qwen3-14B2026.04 | 91.48 | 74.75 | 63.15 | |
| LTEModel=Falcon3-10B2026.04 | 88.26 | 77.71 | 65.14 | |
| EditCoTModel=Qwen3-14B2026.04 | 87.76 | 80.53 | 76.06 | |
| LTEModel=LLaMA3-8B2026.04 | 87.47 | 75.05 | 67.18 | |
| EditCoTModel=LLaMA3-8B2026.04 | 84.64 | 80.8 | 72.15 | |
| SKEMEModel=Qwen3-14B2026.04 | 82.85 | 56.18 | 55.7 | |
| IKEModel=Qwen3-14B2026.04 | 80.38 | 52.63 | 53.37 | |
| EditCoTModel=Falcon3-10B2026.04 | 80.24 | 75.15 | 71.11 | |
| SKEMEModel=Falcon3-10B2026.04 | 78.51 | 65.1 | 58 | |
| SKEMEModel=LLaMA3-8B2026.04 | 77.35 | 70.13 | 60.68 | |
| IKEModel=Falcon3-10B2026.04 | 75.3 | 61.24 | 61.78 | |
| IKEModel=LLaMA3-8B2026.04 | 74.97 | 65.71 | 58.07 | |
| RoMEModel=LLaMA3-8B2026.04 | 0 | 0 | 0 | |
| MEMITModel=LLaMA3-8B2026.04 | 0 | 0 | 0 | |
| PMETModel=LLaMA3-8B2026.04 | 0 | 0 | 0 | |
| GLAMEModel=LLaMA3-8B2026.04 | 0 | 0 | 0 | |
| AlphaEditModel=LLaMA3-8B2026.04 | 0 | 0 | 0 | |
| RoMEModel=Falcon3-10B2026.04 | 0 | 0 | 0 | |
| MEMITModel=Falcon3-10B2026.04 | 0 | 0 | 0 | |
| PMETModel=Falcon3-10B2026.04 | 0 | 0 | 0 | |
| GLAMEModel=Falcon3-10B2026.04 | 0 | 0 | 0 | |
| AlphaEditModel=Falcon3-10B2026.04 | 0 | 0 | 0 | |
| RoMEModel=Qwen3-14B2026.04 | 0 | 0 | 0 | |
| MEMITModel=Qwen3-14B2026.04 | 0 | 0 | 0 | |
| PMETModel=Qwen3-14B2026.04 | 0 | 0 | 0 | |
| GLAMEModel=Qwen3-14B2026.04 | 0 | 0 | 0 | |
| AlphaEditModel=Qwen3-14B2026.04 | 0 | 0 | 0 |