Share your thoughts, 1 month free Claude Pro on usSee more

Model Editing on FEVER 20K edits (test)

99.07Efficacy

STABLEEDIT

Updated 21d ago

Evaluation Results

Method	Links
STABLEEDIT 2026.05		99.07	94.87	69.61
STABLEEDIT 2026.05		98.38	96.55	83.89
STABLEEDIT 2026.05		98.35	96.99	79.85
ULTRAEDIT 2026.05		97.97	93.91	68.86
STABLEEDIT 2026.05		97.89	94.47	69.24
ULTRAEDIT 2026.05		97.87	96.09	84.29
ULTRAEDIT 2026.05		97.45	96.37	79.72
RLEDIT 2026.05		97.29	96.01	79.86
RLEDIT 2026.05		96.93	93.92	68.28
ULTRAEDIT 2026.05		95.39	91.93	67.14
MALMEN 2026.05		95.03	92.44	69.2
RLEDIT 2026.05		92.35	91.39	71.85
RLEDIT 2026.05		91.38	89.93	41.98
MEND 2026.05		76.43	77.66	40.39
MEND 2026.05		52.8	51.44	45.42
MEND 2026.05		36.19	36.19	24.31
ALPHAEDIT 2026.05		32.78	31.19	22.12
FT 2026.05		26.09	24.62	21.36
FT 2026.05		23.8	23.37	16.3
MALMEN 2026.05		18.42	17.43	12.09
FT 2026.05		16.21	13.08	5.01
FT 2026.05		14.02	13.98	9.26
Pre-edited 2026.05		9.61	9.68	15.9
ALPHAEDIT 2026.05		6.39	6.14	2.72
MEMIT 2026.05		5.54	5.03	5.46
ALPHAEDIT 2026.05		1.89	1.87	1.85
MALMEN 2026.05		1.33	1.25	2.92
Pre-edited 2026.05		0.57	0.6	2.17
Pre-edited 2026.05		0.41	0.5	1.98
MEMIT 2026.05		0.08	0.12	0.15
MALMEN 2026.05		0.06	0.06	0.07
Pre-edited 2026.05		0.02	0.02	0.26
MEMIT 2026.05		0.02	0.02	0.02
MEMIT 2026.05		0	0	0
ALPHAEDIT 2026.05		0	0	0
MEND 2026.05		0	0	0