Model Editing on ZsRE (Reliability, Generalization, MMLU, IFEval, TruthfulQA, ARC-C, GSM8K) benchmark leaderboard · SOTA2 Research

Model Editing on ZsRE (Reliability, Generalization, MMLU, IFEval, TruthfulQA, ARC-C, GSM8K)

80.5ReliabilityCrispEdit

Evaluation Results

Method	Links
CrispEdit2026.02		80.5	69	69.5	67.9	50.5	55	76
CrispEdit-Seq2026.02		72.8	60.6	67.8	70.2	53.6	52	74
CrispEdit-Seq2026.02		71.1	62.9	67.8	70.2	53.6	52	74
AlphaEdit2026.02		70.1	60.6	52.7	47.7	46.3	40.5	45.5
LocBF-FT2026.02		69.5	59.7	69.5	70.1	51.6	54	75.5
CrispEdit2026.02		57.4	50.9	69.5	67.9	50.5	55	76
AlphaEdit2026.02		48.1	39.4	52.7	47.7	46.3	40.5	45.5
FT2026.02		46.8	43.1	69.3	45	48.7	43	50
LocBF-FT2026.02		25.2	22.1	69.5	70.1	51.6	54	75.5
UltraEdit2026.02		22.7	17.4	69.3	72.5	51.8	54.5	73
UltraEdit2026.02		20	16.3	69.3	72.5	51.8	54.5	73
LoRA2026.02		18.7	7.2	67.8	70.8	52	56	71
Adam-NSCL2026.02		16.6	15.5	69.2	29.6	50.8	42	39.5
FT2026.02		9.9	8.3	69.3	45	48.7	43	50
LoRA2026.02		9.1	7.4	67.8	70.8	52	56	71
LoRA Sequential2026.02		4.4	4	67.3	64.6	56	47	67
FT Sequential2026.02		3.6	3.5	68.8	19.4	52.8	40.5	6.5
LLaMA-3-8B-Instruct2026.02		2.9	2.1	69.5	69.3	50.7	58	73.5
LLaMA-3-8B-Instruct2026.02		2.1	1.7	69.5	69.3	50.7	58	73.5
Adam-NSCL2026.02		1.9	2	69.2	29.6	50.8	42	39.5
LoRA Sequential2026.02		1.3	0.9	67.3	64.6	56	47	67
FT Sequential2026.02		0.9	1.2	68.8	19.4	52.8	40.5	6.5
MEMIT2026.02		0.1	0	22.9	0	51.3	23.5	0
MEMIT2026.02		0.1	0.1	22.9	0	51.3	23.5	0
MEND2026.02		0	0	22.9	18.2	0	26	0
MEND2026.02		0	0	22.9	18.2	0	26	0