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-accuracy_distance")
score = benchmark(my_model)

Benchmark API

Code examples

Model scores

Score Legend Min Alignment Max Alignment

Rank	Model	Score
1	alexnet_training_seed_03	1.0
2	convnext_xxlarge:clip_laion2b_soup_ft_in1k	.921
3	convnext_xxlarge_clip	.917
4	vit_huge_clip336_laion	.716
5	eva02_large_in22k	.703
6	convnext_large_clip_laion2b_finelayers	.656
7	vit_huge_patch14_clip_336:laion2b_ft_in12k_in1k	.645
8	vit_huge_patch14_clip_224:laion2b_ft_in12k_in1k	.635
9	mehrer2020_alexnet_training_seed_01_fov4	.573
10	vit_large_patch14_clip_224:openai_ft_in1k	.526
11	vit_large_patch14_clip_224:openai_ft_in12k_in1k	.490
12	vit_large_patch14_clip_336:laion2b_ft_in1k	.469
13	convnext_large_mlp:clip_laion2b_augreg_ft_in1k_384	.463
14	vit_large_patch14_clip_224:laion2b_ft_in12k_in1k	.452
15	vit_large_patch14_clip_224:laion2b_ft_in1k	.439
16	dinov2_vitg14_lc	.409
17	vit_large_patch14_clip_336:openai_ft_in12k_in1k	.390
18	dinov2_vitl14_lc	.376
19	vit_large_clip336_openai	.366
20	dinov2_vitg14_reg_lc	.352
21	resnet_50_v1_spiking_l4a	.348
22	alexnet_training_seed_09	.322
23	dinov2_vitl14_reg_lc	.322
24	resnext101_32x48d_wsl	.277
25	ReAlnet04	.276
26	densenet-201	.275
27	resnet50-moclr8deg	.263
28	dinov2_vitl14_lc_res336	.258
29	cvt_cvt-21-384-in1k_4_LucyV4	.257
30	convnext_femto_ols:d1_in1k	.254
31	resnet50-SIN	.254
32	convnext_tiny_sup	.251
33	resnet50_barlowtwins_linhead	.249
34	convnext_large:fb_in22k_ft_in1k	.246
35	convnext_base_fb_in22k_ft_in1k_gram_aligned_dinov3	.232
36	cvt_cvt-21-224-in1k_4_LucyV4	.230
37	convnext_large_in22k_supervised	.227
38	resnet50_ewconplus_wzc_seed5_2__323	.220
39	resnet50_st_wzc_seed3_2_323	.219
40	convnext_base:clip_laiona_augreg_ft_in1k_384	.217
41	convnext_base_imagenet_full_seed-0	.214
42	cvt_cvt-13-384-in22k_finetuned-in1k_4_LucyV4	.202
43	resnet50_st_wzc_seed7_2_323	.202
44	cvt_cvt-w24-384-in22k_finetuned-in1k_4	.201
45	grcnn	.201
46	antialiased-r50	.198
47	resnet50_simsiam_linhead	.191
48	cvt_cvt-13-224-in1k_4_LucyV4	.191
49	convnext_large_imagenet_full_seed-0	.191
50	convnext_xlarge:fb_in22k_ft_in1k	.191
51	grcnn_109	.191
52	pnasnet_large_halfPrecision	.190
53	resnet34_imagenet_full	.189
54	resnet50_imagenet_full	.188
55	deit3_base_patch16_224_fb_in1k	.185
56	resnext101_32x16d_wsl	.184
57	resnet_50_v2	.183
58	resnet_152_v2	.181
59	deit3_base_patch16_224_fb_in22k_ft_in1k_gram_aligned_dinov3	.177
60	convnext_tiny_imagenet_full_seed-0	.176
61	convnext_base_fb_in22k_ft_in1k	.176
62	vit_relpos_base_patch32_plus_rpn_256:sw_in1k	.175
63	cvt_cvt-13-384-in1k_4_LucyV4	.174
64	AdvProp_efficientnet-b6	.173
65	resnet50_st_wzc_seed6_2_323	.172
66	mehrer2020_alexnet_training_seed_04	.171
67	mobilenet_v2_1_4_224	.171
68	resnet50_vicreg_linhead	.169
69	resnet50_ewconplus_wzc_seed2_2__323	.169
70	blt_vs	.167
71	resnet18_imagenet_full	.167
72	ReAlnet09	.165
73	seresnext50_32x4d	.162
74	focalnet_tiny_lrf_in1k	.161
75	resnet50_robust_l2_eps3	.160
76	imagenet_l2_3_0	.160
77	resnet-50-robust	.160
78	resnext101_32x32d_wsl	.157
79	resnet101_imagenet_full	.157
80	resnet_101_v1	.156
81	resnet50-SIN_IN	.152
82	resnet50_ewconplus_wzc_seed6_2_323	.152
83	convnext_tiny:in12k_ft_in1k	.151
84	mehrer2020_alexnet_training_seed_01	.151
85	Res2Net50_26w_4s	.151
86	convnext_small_imagenet_100_seed-0	.149
87	resnet152_imagenet_full	.149
88	resnet_SIN_IN_FT_IN	.148
89	resnet50-SIN_IN_IN	.148
90	convnext_small_imagenet_full_seed-0	.144
91	swin_small_patch4_window7_224:ms_in22k_ft_in1k	.140
92	efficientnet_b0	.140
93	AdvProp_efficientnet-b2	.138
94	AT_efficientnet-b2	.138
95	effnetb1_272x240	.138
96	deit_base_imagenet_full_seed-0	.138
97	mehrer2020_alexnet_training_seed_02	.138
98	voneresnet-50-non_stochastic	.138
99	custom_model_cv_18_dagger_408	.136
100	nasnet_mobile	.136
101	resnet50-vitoimagevidnet8	.136
102	resnet50-VITO-8deg-cc	.136
103	resnet_101_v2	.135
104	AdvProp_efficientnet-b7	.135
105	vit_relpos_base_patch16_clsgap_224:sw_in1k	.133
106	deit3_base_patch16_224_fb_in1k_gram_aligned_dinov3	.131
107	antialias-resnet152	.131
108	ijepa_vith14_linhead	.131
109	ViT_L_32_imagenet1k	.130
110	convnext_base_fb_in1k	.126
111	effnetb1_cutmixpatch_augmix_robust32_avge4e7_manylayers_324x288	.125
112	shufflenet_v2_x1_0	.125
113	voneresnet-50-robust	.125
114	resnext101_32x8d_wsl	.125
115	regnety_032	.124
116	vit_base_patch16_clip_224:openai_ft_in12k_in1k	.124
117	effnetb1_cutmix_augmix_sam_e1_5avg_424x377	.123
118	resnet50_ewconplus_wzc	.121
119	ReAlnet08	.119
120	resnet50_byol_linhead	.115
121	ReAlnet10	.115
122	mehrer2020_alexnet_training_seed_03	.114
123	deit_large_imagenet_full_seed-0	.111
124	resnet_50_v1_spiking	.111
125	mehrer2020_alexnet_training_seed_05	.107
126	vit_tiny_r_s16_p8_384:augreg_in21k_ft_in1k	.107
127	cv_18_dagger_408_pretrained	.107
128	resnet50_st_wzc_2	.104
129	densenet-121	.103
130	convnext_base_fb_in1k_gram_aligned_dinov3	.103
131	deit3_base_patch16_224_fb_in22k_ft_in1k	.103
132	AdvProp_efficientnet-b8	.102
133	resnet_50_v1	.101
134	resnet50_tutorial	.101
135	resnet50-sup	.101
136	vit_base_patch16_clip_224:openai_ft_in1k	.101
137	efficientnet_b2_imagenet_full	.090
138	vonegrcnn_47e	.089
139	resnet50_hvs_wzc	.089
140	resnet50_robust_l2_eps1	.088
141	deit_small_imagenet_full_seed-0	.084
142	AdvProp_efficientnet-b4	.083
143	efficientnet_b1_imagenet_full	.081
144	ReAlnet07	.080
145	voneresnet-50	.077
146	cvt_cvt-21-384-in22k_finetuned-in1k_4_LucyV4	.071
147	vonegrcnn_62e_nobn	.070
148	resnet50_simclr_linhead_01norm	.069
149	efficientnet_b0_imagenet_full	.066
150	mehrer2020_alexnet_training_seed_01_fov12	.064
151	inception_v3	.064
152	effnetb1_cutmixpatch_SAM_robust32_avge6e8e9e10_manylayers_324x288	.058
153	eicircuit_resnet18_bn_20260919	.055
154	alexnet	.052
155	AlexNet_SIN_fov12	.043
156	AlexNet_SIN	.043
157	mehrer2020_alexnet_training_seed_01_fov16	.000
158	vgg_16	.000
159	alexnet_training_seed_07	.000
160	AlexNet_SIN_fov16	.000
161	mobilenet_v2_1_0_128
162	gabor_filter_entropy_4x4_v2_forcedCosine
163	gabor_filter_entropy_4x4_v2
164	dinov2_vitl14_reg_lc_featread
165	deepeisnn_ei_imagenet_resnet18
166	alexnet_training_seed_01
167	alexnet_training_seed_10
168	alexnet_training_seed_06
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

0.78.

Note that scores are relative to this ceiling.

Data: Kato2026Sparse

Metric: accuracy_distance