Model Editing on ZsRE (3,000 samples, test)
99.1Relational ScoreLocBF-FT
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| LocBF-FTbackbone=LLaMA-3-8B-Instruct, pipeline=teacher-forcing2026.02 | 99.1 | 91.1 | 34.5 | |
| FTbackbone=LLaMA-3-8B-Instruct, pipeline=teacher-forcing2026.02 | 99.1 | 93.1 | 22.9 | |
| CrispEditbackbone=LLaMA-3-8B-Instruct, pipeline=teacher-forcing2026.02 | 99.1 | 92.1 | 32.3 | |
| Adam-NSCLbackbone=LLaMA-3-8B-Instruct, pipeline=teacher-forcing2026.02 | 98.8 | 92.4 | 22.1 | |
| CrispEdit-Seqbackbone=LLaMA-3-8B-Instruct, pipeline=teacher-forcing, editing_mode=sequential2026.02 | 98.3 | 91.4 | 30.2 | |
| LoRAbackbone=LLaMA-3-8B-Instruct, pipeline=teacher-forcing2026.02 | 93.4 | 60.6 | 30.9 | |
| AlphaEditbackbone=LLaMA-3-8B-Instruct, pipeline=teacher-forcing2026.02 | 86.7 | 77.8 | 32.4 | |
| FT Sequentialbackbone=LLaMA-3-8B-Instruct, pipeline=teacher-forcing, editing_mode=sequential2026.02 | 79.7 | 76.6 | 16.8 | |
| UltraEditbackbone=LLaMA-3-8B-Instruct, pipeline=teacher-forcing2026.02 | 61.9 | 57.3 | 31.5 | |
| LoRA Sequentialbackbone=LLaMA-3-8B-Instruct, pipeline=teacher-forcing, editing_mode=sequential2026.02 | 36.8 | 32.7 | 21.7 | |
| Llama 3 8B Instructbackbone=LLaMA-3-8B-Instruct, pipeline=teacher-forcing2026.02 | 25.7 | 25.1 | 37.8 | |
| MEMITbackbone=LLaMA-3-8B-Instruct, pipeline=teacher-forcing2026.02 | 0 | 0 | 0 | |
| MENDbackbone=LLaMA-3-8B-Instruct, pipeline=teacher-forcing2026.02 | 0 | 0 | 0.1 |