Leaderboard Explore Compare Profile Community Blog FAQ Competition 2024 Report Issue

Sample stimuli

How to use

from brainscore_vision import load_benchmark
benchmark = load_benchmark("Kato2026Sparse-error_consistency")
score = benchmark(my_model)

Benchmark API

Code examples

Model scores

Score Legend Min Alignment Max Alignment

Rank	Model	Score
1	convnext_xxlarge:clip_laion2b_soup_ft_in1k	.831
2	convnext_xxlarge_clip	.667
3	vit_huge_clip336_laion	.641
4	vit_huge_patch14_clip_224:laion2b_ft_in12k_in1k	.575
5	vit_huge_patch14_clip_336:laion2b_ft_in12k_in1k	.548
6	eva02_large_in22k	.459
7	vit_large_patch14_clip_224:openai_ft_in1k	.442
8	dinov2_vitg14_lc	.401
9	convnext_large_clip_laion2b_finelayers	.368
10	dinov2_vitl14_lc	.331
11	vit_large_patch14_clip_224:openai_ft_in12k_in1k	.315
12	resnet_50_v1_spiking_l4a	.311
13	vit_large_patch14_clip_336:openai_ft_in12k_in1k	.292
14	vit_large_patch14_clip_224:laion2b_ft_in12k_in1k	.240
15	vit_large_patch14_clip_224:laion2b_ft_in1k	.215
16	vit_large_patch14_clip_336:laion2b_ft_in1k	.181
17	dinov2_vitl14_reg_lc	.175
18	dinov2_vitg14_reg_lc	.175
19	vit_large_clip336_openai	.173
20	dinov2_vitl14_lc_res336	.149
21	convnext_large_mlp:clip_laion2b_augreg_ft_in1k_384	.120
22	resnet50-SIN	.102
23	convnext_large:fb_in22k_ft_in1k	.100
24	ReAlnet04	.081
25	densenet-201	.037
26	convnext_xlarge:fb_in22k_ft_in1k	.035
27	AdvProp_efficientnet-b7	.034
28	deit3_base_patch16_224_fb_in22k_ft_in1k_gram_aligned_dinov3	.028
29	resnext101_32x48d_wsl	.027
30	convnext_base:clip_laiona_augreg_ft_in1k_384	.018
31	convnext_base_fb_in22k_ft_in1k	.018
32	convnext_large_imagenet_full_seed-0	.015
33	convnext_large_in22k_supervised	.013
34	vit_base_patch16_clip_224:openai_ft_in12k_in1k	.002
35	voneresnet-50	.000
36	cvt_cvt-w24-384-in22k_finetuned-in1k_4	.000
37	deit_large_imagenet_full_seed-0	.000
38	convnext_femto_ols:d1_in1k	.000
39	convnext_base_imagenet_full_seed-0	.000
40	vit_relpos_base_patch16_clsgap_224:sw_in1k	.000
41	resnet50_ewconplus_wzc_seed2_2__323	.000
42	resnet50-SIN_IN	.000
43	resnet50_byol_linhead	.000
44	Res2Net50_26w_4s	.000
45	mehrer2020_alexnet_training_seed_05	.000
46	mehrer2020_alexnet_training_seed_01_fov12	.000
47	AdvProp_efficientnet-b4	.000
48	resnet_101_v2	.000
49	nasnet_mobile	.000
50	cvt_cvt-21-384-in1k_4_LucyV4	.000
51	resnet_101_v1	.000
52	vonegrcnn_62e_nobn	.000
53	convnext_base_fb_in1k	.000
54	efficientnet_b2_imagenet_full	.000
55	resnet50_robust_l2_eps1	.000
56	resnet50_ewconplus_wzc_seed6_2_323	.000
57	resnet50_st_wzc_seed7_2_323	.000
58	resnet50_simclr_linhead_01norm	.000
59	convnext_tiny:in12k_ft_in1k	.000
60	cvt_cvt-13-384-in1k_4_LucyV4	.000
61	resnext101_32x16d_wsl	.000
62	resnet50-vitoimagevidnet8	.000
63	resnet50-VITO-8deg-cc	.000
64	resnext101_32x32d_wsl	.000
65	deit_base_imagenet_full_seed-0	.000
66	efficientnet_b0_imagenet_full	.000
67	resnet50_st_wzc_seed3_2_323	.000
68	resnet_50_v1	.000
69	resnet50_tutorial	.000
70	resnet50-sup	.000
71	ReAlnet08	.000
72	vit_base_patch16_clip_224:openai_ft_in1k	.000
73	resnet50_st_wzc_2	.000
74	deit3_base_patch16_224_fb_in22k_ft_in1k	.000
75	alexnet	.000
76	cvt_cvt-21-384-in22k_finetuned-in1k_4_LucyV4	.000
77	resnet50_hvs_wzc	.000
78	ReAlnet09	.000
79	eicircuit_resnet18_bn_20260919	.000
80	resnet50_barlowtwins_linhead	.000
81	ijepa_vith14_linhead	.000
82	deit_small_imagenet_full_seed-0	.000
83	resnext101_32x8d_wsl	.000
84	resnet_152_v2	.000
85	AdvProp_efficientnet-b8	.000
86	seresnext50_32x4d	.000
87	resnet_50_v1_spiking	.000
88	vit_tiny_r_s16_p8_384:augreg_in21k_ft_in1k	.000
89	vonegrcnn_47e	.000
90	resnet34_imagenet_full	.000
91	mehrer2020_alexnet_training_seed_01_fov16	.000
92	mehrer2020_alexnet_training_seed_03	.000
93	inception_v3	.000
94	mehrer2020_alexnet_training_seed_02	.000
95	densenet-121	.000
96	efficientnet_b1_imagenet_full	.000
97	antialias-resnet152	.000
98	AdvProp_efficientnet-b6	.000
99	cvt_cvt-21-224-in1k_4_LucyV4	.000
100	convnext_base_fb_in22k_ft_in1k_gram_aligned_dinov3	.000
101	mehrer2020_alexnet_training_seed_04	.000
102	deit3_base_patch16_224_fb_in1k	.000
103	convnext_base_fb_in1k_gram_aligned_dinov3	.000
104	shufflenet_v2_x1_0	.000
105	convnext_tiny_sup	.000
106	regnety_032	.000
107	antialiased-r50	.000
108	resnet50_vicreg_linhead	.000
109	voneresnet-50-non_stochastic	.000
110	focalnet_tiny_lrf_in1k	.000
111	efficientnet_b0	.000
112	mehrer2020_alexnet_training_seed_01	.000
113	cv_18_dagger_408_pretrained	.000
114	resnet50_st_wzc_seed6_2_323	.000
115	ViT_L_32_imagenet1k	.000
116	resnet50_simsiam_linhead	.000
117	deit3_base_patch16_224_fb_in1k_gram_aligned_dinov3	.000
118	effnetb1_272x240	.000
119	ReAlnet07	.000
120	effnetb1_cutmixpatch_augmix_robust32_avge4e7_manylayers_324x288	.000
121	AT_efficientnet-b2	.000
122	AdvProp_efficientnet-b2	.000
123	resnet_50_v2	.000
124	AlexNet_SIN	.000
125	resnet152_imagenet_full	.000
126	resnet50_ewconplus_wzc_seed5_2__323	.000
127	resnet50-moclr8deg	.000
128	convnext_small_imagenet_full_seed-0	.000
129	resnet50_ewconplus_wzc	.000
130	resnet18_imagenet_full	.000
131	AlexNet_SIN_fov16	.000
132	convnext_tiny_imagenet_full_seed-0	.000
133	resnet50_imagenet_full	.000
134	resnet_SIN_IN_FT_IN	.000
135	resnet50-SIN_IN_IN	.000
136	blt_vs	.000
137	ReAlnet10	.000
138	pnasnet_large_halfPrecision	.000
139	grcnn	.000
140	AlexNet_SIN_fov12	.000
141	effnetb1_cutmix_augmix_sam_e1_5avg_424x377	.000
142	grcnn_109	.000
143	vit_relpos_base_patch32_plus_rpn_256:sw_in1k	.000
144	mobilenet_v2_1_4_224	.000
145	custom_model_cv_18_dagger_408	.000
146	vgg_16	.000
147	voneresnet-50-robust	.000
148	effnetb1_cutmixpatch_SAM_robust32_avge6e8e9e10_manylayers_324x288	.000
149	convnext_small_imagenet_100_seed-0	.000
150	cvt_cvt-13-224-in1k_4_LucyV4	.000
151	cvt_cvt-13-384-in22k_finetuned-in1k_4_LucyV4	.000
152	swin_small_patch4_window7_224:ms_in22k_ft_in1k	.000
153	resnet101_imagenet_full	.000
154	resnet50_robust_l2_eps3	.000
155	imagenet_l2_3_0	.000
156	resnet-50-robust	.000
157	mehrer2020_alexnet_training_seed_01_fov4
158	mobilenet_v2_1_0_128
159	gabor_filter_entropy_4x4_v2_forcedCosine
160	gabor_filter_entropy_4x4_v2
161	dinov2_vitl14_reg_lc_featread
162	deepeisnn_ei_imagenet_resnet18
163	alexnet_training_seed_01
164	alexnet_training_seed_10
165	alexnet_training_seed_09
166	alexnet_training_seed_07
167	alexnet_training_seed_06
168	alexnet_training_seed_03
169	alexnet_training_seed_02
170	hmax
171	ReAlnet01_cornet
172	compact_resnet50_robust_V4
173	VOneCORnet-S
174	ReAlnet04_cornet
175	ReAlnet03_cornet
176	ReAlnet01_cornet_fov4

Benchmark bibtex

None

Ceiling

0.39.

Note that scores are relative to this ceiling.

Data: Kato2026Sparse

Metric: error_consistency