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("Kato2026Ori-error_consistency")
score = benchmark(my_model)

Benchmark API

Code examples

Model scores

Score Legend Min Alignment Max Alignment

Rank	Model	Score
1	resnext101_32x32d_wsl	.734
2	vit_large_patch14_clip_336:laion2b_ft_in1k	.716
3	AdvProp_efficientnet-b7	.674
4	vit_base_patch16_clip_224:openai_ft_in1k	.667
5	convnext_small_imagenet_full_seed-0	.660
6	convnext_base_imagenet_full_seed-0	.658
7	deit3_base_patch16_224_fb_in1k	.641
8	pnasnet_large_halfPrecision	.635
9	vit_large_clip336_openai	.630
10	vit_large_patch14_clip_336:openai_ft_in12k_in1k	.630
11	vit_large_patch14_clip_224:openai_ft_in1k	.630
12	effnetb1_272x240	.623
13	voneresnet-50-non_stochastic	.614
14	deit_small_imagenet_full_seed-0	.596
15	focalnet_tiny_lrf_in1k	.582
16	deit_large_imagenet_full_seed-0	.579
17	convnext_large_imagenet_full_seed-0	.569
18	resnet152_imagenet_full	.569
19	seresnext50_32x4d	.568
20	cvt_cvt-13-384-in1k_4_LucyV4	.553
21	convnext_base_fb_in1k	.547
22	cvt_cvt-21-384-in22k_finetuned-in1k_4_LucyV4	.547
23	Res2Net50_26w_4s	.536
24	cvt_cvt-13-384-in22k_finetuned-in1k_4_LucyV4	.535
25	cvt_cvt-w24-384-in22k_finetuned-in1k_4	.530
26	resnext101_32x16d_wsl	.518
27	convnext_large_in22k_supervised	.517
28	deit3_base_patch16_224_fb_in22k_ft_in1k	.514
29	resnet50_hvs_wzc	.514
30	vonegrcnn_62e_nobn	.506
31	antialias-resnet152	.496
32	densenet-201	.493
33	resnet50_robust_l2_eps1	.492
34	convnext_femto_ols:d1_in1k	.488
35	convnext_xxlarge_clip	.488
36	mobilenet_v2_1_4_224	.474
37	eva02_large_in22k	.474
38	swin_small_patch4_window7_224:ms_in22k_ft_in1k	.469
39	grcnn_109	.467
40	cvt_cvt-13-224-in1k_4_LucyV4	.459
41	vit_large_patch14_clip_224:laion2b_ft_in12k_in1k	.458
42	dinov2_vitl14_lc	.452
43	ijepa_vith14_linhead	.448
44	convnext_large:fb_in22k_ft_in1k	.442
45	grcnn	.440
46	inception_v3	.440
47	vgg_16	.435
48	resnet50_ewconplus_wzc_seed5_2__323	.433
49	resnet_50_v2	.431
50	resnet_101_v1	.421
51	vit_huge_clip336_laion	.420
52	resnet50_ewconplus_wzc_seed6_2_323	.419
53	deit3_base_patch16_224_fb_in22k_ft_in1k_gram_aligned_dinov3	.417
54	resnext101_32x8d_wsl	.415
55	dinov2_vitl14_lc_res336	.411
56	convnext_tiny_imagenet_full_seed-0	.409
57	resnet50_barlowtwins_linhead	.409
58	ReAlnet04	.405
59	convnext_xlarge:fb_in22k_ft_in1k	.405
60	resnet50_st_wzc_seed6_2_323	.403
61	resnet50_st_wzc_2	.403
62	resnext101_32x48d_wsl	.401
63	dinov2_vitl14_reg_lc	.400
64	resnet34_imagenet_full	.399
65	AdvProp_efficientnet-b4	.394
66	ViT_L_32_imagenet1k	.393
67	resnet50_ewconplus_wzc	.392
68	efficientnet_b0	.389
69	ReAlnet07	.385
70	resnet50_imagenet_full	.383
71	antialiased-r50	.379
72	vit_large_patch14_clip_224:laion2b_ft_in1k	.379
73	cvt_cvt-21-224-in1k_4_LucyV4	.376
74	resnet50_simsiam_linhead	.376
75	resnet101_imagenet_full	.375
76	resnet50_ewconplus_wzc_seed2_2__323	.373
77	AdvProp_efficientnet-b6	.369
78	vit_huge_patch14_clip_224:laion2b_ft_in12k_in1k	.368
79	resnet50_st_wzc_seed7_2_323	.364
80	resnet_50_v1_spiking	.361
81	convnext_xxlarge:clip_laion2b_soup_ft_in1k	.360
82	ReAlnet09	.359
83	dinov2_vitg14_lc	.358
84	ReAlnet08	.356
85	voneresnet-50	.352
86	convnext_base_fb_in22k_ft_in1k	.351
87	ReAlnet10	.342
88	resnet_50_v1	.341
89	resnet50_tutorial	.341
90	resnet50-sup	.341
91	vit_large_patch14_clip_224:openai_ft_in12k_in1k	.341
92	mehrer2020_alexnet_training_seed_01	.340
93	dinov2_vitg14_reg_lc	.339
94	regnety_032	.333
95	convnext_base_fb_in22k_ft_in1k_gram_aligned_dinov3	.332
96	eicircuit_resnet18_bn_20260919	.332
97	resnet50_st_wzc_seed3_2_323	.329
98	AT_efficientnet-b2	.328
99	AdvProp_efficientnet-b2	.328
100	convnext_large_mlp:clip_laion2b_augreg_ft_in1k_384	.325
101	cv_18_dagger_408_pretrained	.319
102	convnext_tiny_sup	.317
103	AdvProp_efficientnet-b8	.316
104	resnet50-SIN_IN	.315
105	vit_huge_patch14_clip_336:laion2b_ft_in12k_in1k	.313
106	vit_relpos_base_patch16_clsgap_224:sw_in1k	.308
107	convnext_tiny:in12k_ft_in1k	.307
108	convnext_base:clip_laiona_augreg_ft_in1k_384	.306
109	vit_relpos_base_patch32_plus_rpn_256:sw_in1k	.300
110	shufflenet_v2_x1_0	.297
111	convnext_large_clip_laion2b_finelayers	.284
112	blt_vs	.283
113	nasnet_mobile	.282
114	resnet18_imagenet_full	.275
115	resnet50_vicreg_linhead	.267
116	convnext_base_fb_in1k_gram_aligned_dinov3	.267
117	mehrer2020_alexnet_training_seed_03	.265
118	mehrer2020_alexnet_training_seed_05	.265
119	vit_base_patch16_clip_224:openai_ft_in12k_in1k	.263
120	mehrer2020_alexnet_training_seed_01_fov12	.260
121	deit3_base_patch16_224_fb_in1k_gram_aligned_dinov3	.258
122	efficientnet_b2_imagenet_full	.254
123	densenet-121	.247
124	voneresnet-50-robust	.246
125	deit_base_imagenet_full_seed-0	.242
126	resnet_152_v2	.235
127	resnet_101_v2	.220
128	efficientnet_b1_imagenet_full	.217
129	resnet50-moclr8deg	.213
130	mehrer2020_alexnet_training_seed_02	.209
131	resnet50-SIN	.208
132	custom_model_cv_18_dagger_408	.208
133	alexnet	.207
134	vit_tiny_r_s16_p8_384:augreg_in21k_ft_in1k	.203
135	resnet_SIN_IN_FT_IN	.197
136	resnet50-SIN_IN_IN	.197
137	cvt_cvt-21-384-in1k_4_LucyV4	.186
138	AlexNet_SIN	.185
139	resnet50-vitoimagevidnet8	.174
140	resnet50-VITO-8deg-cc	.174
141	vonegrcnn_47e	.166
142	efficientnet_b0_imagenet_full	.157
143	convnext_small_imagenet_100_seed-0	.153
144	resnet_50_v1_spiking_l4a	.143
145	resnet50_byol_linhead	.133
146	resnet50_simclr_linhead_01norm	.111
147	mehrer2020_alexnet_training_seed_04	.097
148	mehrer2020_alexnet_training_seed_01_fov16	.086
149	resnet50_robust_l2_eps3	.073
150	imagenet_l2_3_0	.073
151	resnet-50-robust	.073
152	mehrer2020_alexnet_training_seed_01_fov4	.071
153	effnetb1_cutmix_augmix_sam_e1_5avg_424x377	.034
154	effnetb1_cutmixpatch_SAM_robust32_avge6e8e9e10_manylayers_324x288	.025
155	AlexNet_SIN_fov12	.023
156	effnetb1_cutmixpatch_augmix_robust32_avge4e7_manylayers_324x288	.021
157	AlexNet_SIN_fov16	.012
158	alexnet_training_seed_07	.005
159	alexnet_training_seed_09	.003
160	alexnet_training_seed_03	.003
161	alexnet_training_seed_10	.000
162	alexnet_training_seed_06	.000
163	alexnet_training_seed_02	.000
164	alexnet_training_seed_01	.000
165	mobilenet_v2_1_0_128
166	gabor_filter_entropy_4x4_v2_forcedCosine
167	gabor_filter_entropy_4x4_v2
168	deepeisnn_ei_imagenet_resnet18
169	hmax
170	ReAlnet03_cornet
171	dinov2_vitl14_reg_lc_featread
172	ReAlnet01_cornet
173	compact_resnet50_robust_V4
174	VOneCORnet-S
175	ReAlnet04_cornet
176	ReAlnet01_cornet_fov4

Benchmark bibtex

None

Ceiling

0.21.

Note that scores are relative to this ceiling.

Data: Kato2026Ori

Metric: error_consistency