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

Online Multimodal Model Editing on E-VQA

100Reliability

FT-L

Updated 22d ago

Evaluation Results

Method
FT-L 2026.05	100	100	60	94.74	100	90.95
FT-M 2026.05	100	96.67	63.33	100	73.33	86.67
MEND 2026.05	100	100	100	60.61	33.33	78.79
SERAC 2026.05	100	100	100	89.47	80	93.89
IKE 2026.05	100	100	100	52.63	13.33	73.19
M-ORE 2026.05	100	100	100	94.74	100	98.95
M-ORE 2026.05	100	100	85.12	100	100	97.02
FT-L 2026.07	100	100	60	94.74	100	90.95
FT-M 2026.07	100	96.67	63.33	100	73.33	86.67
MEND 2026.07	100	100	100	60.61	33.33	78.79
M-ORE 2026.07	100	100	96.67	100	100	99.45
SERAC 2026.07	100	100	100	89.47	80	93.89
IKE 2026.07	100	100	100	52.63	13.33	73.19
ScopeEdit 2026.07	100	100	100	100	100	100
M-ORE 2026.07	100	100	84.8	100	100	97.47
ScopeEdit 2026.07	100	100	100	100	100	100
M-ORE 2026.07	99.13	90.34	83.91	99.29	96.5	93.77
ScopeEdit 2026.07	98.8	91.84	88.84	98.94	96.8	95.11
M-ORE 2026.05	97.43	94.3	85.4	96.32	91.9	93.07
M-ORE 2026.05	96.05	92.05	94.06	91.77	88.85	92.56
FT-L 2026.05	95.66	99	81.84	89.38	89.98	91.17
FT-L 2026.07	95.66	99	81.84	89.38	89.98	91.17
MEND 2026.05	95.53	95.53	83.79	74.82	59.65	81.86
MEND 2026.07	95.53	95.53	83.79	74.82	59.65	81.86
FT-M 2026.05	95	95	79.67	100	85.83	91.1
FT-M 2026.07	95	95	79.67	100	85.83	91.1
M-ORE 2026.07	93.51	90.15	80.76	96.08	89.98	90.1
LiveEdit 2026.05	93.36	93.67	87.91	100	100	94.99
LiveEdit 2026.07	93.36	93.67	87.91	100	100	94.99
LiveEdit 2026.05	92.68	92.33	89.25	100	95.57	93.97
LiveEdit 2026.07	92.68	92.33	89.25	100	95.57	93.97
ScopeEdit 2026.07	92.52	96.05	84.18	98.69	91.39	92.57
LiveEdit 2026.05	91.83	91.16	85.02	99.31	92.78	92.02
LiveEdit 2026.07	91.83	91.16	85.02	99.31	92.78	92.02
LiveEdit 2026.05	90.22	91.39	81.49	98.05	95.27	91.28
LiveEdit 2026.07	90.22	91.39	81.49	98.05	95.27	91.28
SERAC 2026.05	90	90	60	100	23.33	72.67
SERAC 2026.07	90	90	60	100	23.33	72.67
SERAC 2026.05	88.03	85.18	88.03	89.95	37.6	77.76
SERAC 2026.07	88.03	85.18	88.03	89.95	37.6	77.76
SERAC 2026.05	87.71	85.03	67.13	92.27	20.83	70.59
SERAC 2026.07	87.71	85.03	67.13	92.27	20.83	70.59
FT-M 2026.05	85.67	81.75	67.03	100	46.19	76.13
FT-M 2026.07	85.67	81.75	67.03	100	46.19	76.13
IKE 2026.05	83.53	83	84.72	74.63	6.37	66.45
IKE 2026.07	83.53	83	84.72	74.63	6.37	66.45
FT-L 2026.05	74.55	66.59	66.14	84.15	65.71	71.43
FT-L 2026.07	74.55	66.59	66.14	84.15	65.71	71.43
AlphaEdit 2026.05	72.57	72.57	70.67	88.04	96.67	80.1
AlphaEdit 2026.07	72.57	72.57	70.67	88.04	96.67	80.1
AlphaEdit 2026.05	71.33	70.93	70.67	88.45	77.67	75.81
AlphaEdit 2026.07	71.33	70.93	70.67	88.45	77.67	75.81
AlphaEdit 2026.05	60	50	40	91.64	73.33	62.99
AlphaEdit 2026.07	60	50	40	91.64	73.33	62.99
FT-M 2026.05	58.4	50.85	43.69	100	46.85	59.96
FT-M 2026.07	58.4	50.85	43.69	100	46.85	59.96
IKE 2026.05	50	50	50	58.33	15	44.67
IKE 2026.07	50	50	50	58.33	15	44.67
IKE 2026.05	38.87	35.85	39.4	46.22	11.24	34.32
IKE 2026.07	38.87	35.85	39.4	46.22	11.24	34.32
AlphaEdit 2026.05	36.83	28.5	26.78	86.35	69.97	49.69
AlphaEdit 2026.07	36.83	28.5	26.78	86.35	69.97	49.69
FT-L 2026.05	26	18	11	86.28	53.12	38.88
FT-L 2026.07	26	18	11	86.28	53.12	38.88
MEND 2026.05	1.09	1.09	1.01	75.33	61.67	28.04
MEND 2026.07	1.09	1.09	1.01	75.33	61.67	28.04
MEND 2026.05	1	1	1	90.42	77.2	34.12
MEND 2026.07	1	1	1	90.42	77.2	34.12