Knowledge Editing on in-prompt benchmark leaderboard · SOTA2 Research

Knowledge Editing on in-prompt

1ReliabilityFT

Evaluation Results

Method	Links
FT2025.07		1	0.717	0.654	0.897	0.589	0.361
FT2025.07		1	0.27	0.668	0.886	0.703	0.404
LoRA2025.07		1	0.268	0.626	0.959	0.462	0.292
FT2025.07		1	0.595	0.579	0.947	0.371	0.274
FT2025.07		1	0.225	0.576	0.962	0.419	0.307
LoRA2025.07		1	0.575	0.502	0.993	0.346	0.264
LoRA2025.07		1	0.176	0.395	0.991	0.291	0.194
LoRA2025.07		0.997	0.693	0.684	0.945	0.491	0.34
MEMIT-Merge2025.07		0.991	0.336	0.644	0.979	0.388	0.299
MEMIT2025.07		0.99	0.321	0.681	0.849	0.397	0.352
MEMIT-Merge2025.07		0.99	0.331	0.684	0.838	0.409	0.357
MEMIT2025.07		0.989	0.356	0.624	0.978	0.389	0.292
MEMIT-Merge2025.07		0.977	0.676	0.699	0.818	0.394	0.341
MEMIT-Merge2025.07		0.961	0.711	0.661	0.964	0.374	0.293
MEMIT2025.07		0.863	0.595	0.646	0.802	0.422	0.381
MEMIT2025.07		0.855	0.713	0.668	0.96	0.428	0.333