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("Zerbe2026_fmri.IT-ood-ridgecv")
score = benchmark(my_model)

Benchmark API

Code examples

Model scores

Score Legend Min Alignment Max Alignment

Rank	Model	Score
1	resnext101_32x48d_wsl	.267
2	resnext101_32x32d_wsl	.263
3	resnext101_32x16d_wsl	.254
4	vit_relpos_base_patch16_clsgap_224:sw_in1k	.251
5	convnext_tiny:in12k_ft_in1k	.243
6	vit_large_patch14_clip_224:laion2b_ft_in12k_in1k	.243
7	dinov2_vitg14_reg_lc	.241
8	deit_small_imagenet_full_seed-0	.244
9	resnet50-moclr8deg	.239
10	deit_large_imagenet_full_seed-0	.242
11	convnext_xxlarge:clip_laion2b_soup_ft_in1k	.235
12	densenet-121	.235
13	pnasnet_large_halfPrecision	.235
14	deit_base_imagenet_full_seed-0	.238
15	ReAlnet08_cornet	.236
16	ReAlnet01_cornet	.232
17	dinov2_vitl14_lc	.231
18	convnext_large:fb_in22k_ft_in1k	.231
19	ReAlnet04_cornet	.230
20	ReAlnet05_cornet	.234
21	convnext_tiny_sup	.230
22	focalnet_tiny_lrf_in1k	.230
23	convnext_base_imagenet_full_seed-0	.230
24	ReAlnet03_cornet	.229
25	ReAlnet06_cornet	.232
26	ReAlnet07_cornet	.232
27	ReAlnet10_cornet	.232
28	ReAlnet02_cornet	.232
29	resnet_50_v2	.229
30	convnext_large_imagenet_full_seed-0	.228
31	convnext_base:clip_laiona_augreg_ft_in1k_384	.228
32	Res2Net50_26w_4s	.228
33	ReAlnet09_cornet	.231
34	resnet50-SIN_IN_IN	.227
35	dinov2_vitl14_lc_res336	.227
36	vit_large_patch14_clip_336:laion2b_ft_in1k	.225
37	VOneCORnet-S	.224
38	resnet50-SIN_IN	.223
39	resnet152_imagenet_full	.222
40	vit_base_patch16_clip_224:openai_ft_in12k_in1k	.223
41	resnet_50_v1_spiking	.222
42	resnet_50_v1	.222
43	resnet50-sup	.222
44	antialiased-r50	.222
45	resnet50_ewconplus_wzc	.222
46	vit_large_patch14_clip_336:openai_ft_in12k_in1k	.221
47	voneresnet-50	.221
48	eva02_large_in22k	.222
49	resnet50_tutorial	.221
50	seresnext50_32x4d	.224
51	regnety_032	.223
52	vit_base_patch16_clip_224:openai_ft_in1k	.220
53	vit_large_patch14_clip_224:openai_ft_in1k	.220
54	ReAlnet01_cornet_fov16	.221
55	resnet_SIN_IN_FT_IN	.217
56	dinov2_vitl14_reg_lc_featread	.216
57	dinov2_vitl14_reg_lc	.216
58	convnext_femto_ols:d1_in1k	.215
59	convnext_small_imagenet_full_seed-0	.215
60	vit_tiny_r_s16_p8_384:augreg_in21k_ft_in1k	.214
61	voneresnet-50-non_stochastic	.213
62	resnet152_ecoset_full	.216
63	custom_model_cv_18_dagger_408	.213
64	resnet50_simclr_linhead_01norm	.215
65	shufflenet_v2_x1_0	.212
66	convnext_tiny_imagenet_full_seed-0	.212
67	efficientnet_b1_imagenet_full	.214
68	vit_large_patch14_clip_224:openai_ft_in12k_in1k	.211
69	resnet50-vitoimagevidnet8	.211
70	resnet50-VITO-8deg-cc	.211
71	resnet50_simsiam_linhead	.213
72	resnet50_byol_linhead	.213
73	effnetb1_cutmix_augmix_sam_e1_5avg_424x377	.210
74	resnet101_ecoset_full	.213
75	resnet50_robust_l2_eps1	.210
76	resnext101_32x8d_wsl	.210
77	convnext_large_mlp:clip_laion2b_augreg_ft_in1k_384	.209
78	eicircuit_resnet18_bn_20260919	.212
79	efficientnet_b0_imagenet_full	.211
80	effnetb1_272x240	.208
81	resnet50_imagenet_full	.208
82	cvt_cvt-w24-384-in22k_finetuned-in1k_4	.208
83	convnext_xxlarge_clip	.207
84	resnet50_ecoset_full	.208
85	vit_huge_clip336_laion	.205
86	efficientnet_b2_imagenet_full	.204
87	ijepa_vith14_linhead	.207
88	resnet101_imagenet_full	.207
89	vit_relpos_base_patch32_plus_rpn_256:sw_in1k	.203
90	resnet50_barlowtwins_linhead	.207
91	mobilenet_v2_1_4_224	.203
92	efficientnet_b0	.202
93	resnet18_imagenet_full	.205
94	resnet50_vicreg_linhead	.204
95	vit_large_patch14_clip_224:laion2b_ft_in1k	.200
96	cvt_cvt-13-224-in1k_4_LucyV4	.199
97	resnet18_ecoset_full	.199
98	swin_small_patch4_window7_224:ms_in22k_ft_in1k	.193
99	resnet34_imagenet_full	.195
100	cvt_cvt-21-224-in1k_4_LucyV4	.191
101	vit_huge_patch14_clip_224:laion2b_ft_in12k_in1k	.190
102	AT_efficientnet-b2	.189
103	AdvProp_efficientnet-b2	.189
104	convnext_xlarge:fb_in22k_ft_in1k	.188
105	cvt_cvt-21-384-in1k_4_LucyV4	.188
106	dinov2_vitg14_lc	.187
107	alexnet	.184
108	AdvProp_efficientnet-b7	.182
109	cvt_cvt-13-384-in22k_finetuned-in1k_4_LucyV4	.183
110	resnet50-SIN	.182
111	ReAlnet01_cornet_fov4	.181
112	effnetb1_cutmixpatch_SAM_robust32_avge6e8e9e10_manylayers_324x288	.180
113	AdvProp_efficientnet-b8	.177
114	cvt_cvt-13-384-in1k_4_LucyV4	.177
115	cvt_cvt-21-384-in22k_finetuned-in1k_4_LucyV4	.175
116	AdvProp_efficientnet-b4	.174
117	AdvProp_efficientnet-b6	.173
118	effnetb1_cutmixpatch_augmix_robust32_avge4e7_manylayers_324x288	.171
119	resnet-50-robust	.169
120	vit_large_clip336_openai	.171
121	ViT_L_32_imagenet1k	.167
122	grcnn	.166
123	mehrer2020_alexnet_training_seed_03	.157
124	mehrer2020_alexnet_training_seed_07	.150
125	mehrer2020_alexnet_training_seed_09	.148
126	mehrer2020_alexnet_training_seed_08	.144
127	mehrer2020_alexnet_training_seed_01_fov4	.142
128	mehrer2020_alexnet_training_seed_10	.142
129	convnext_small_imagenet_100_seed-0	.142
130	mehrer2020_alexnet_training_seed_06	.140
131	deit_small_imagenet_100_seed-0	.133
132	mehrer2020_alexnet_training_seed_01	.133
133	mehrer2020_alexnet_training_seed_04	.131
134	AlexNet_SIN	.119
135	AlexNet_SIN_fov12	.098
136	AlexNet_SIN_fov4	.087
137	mehrer2020_alexnet_training_seed_01_fov16	.085
138	resnet50_imagenet_100_seed-0	.085
139	hmax	.074
140	AlexNet_SIN_fov16	.072
141	gabor_filter_entropy_4x4	.060
142	gabor_filter_entropy_4x4_v2	.052
143	gabor_filter_entropy_4x4_v2_forcedCosine	.048
144	gabor_filter_mean_abs_4x4_v2	.023
145	alexnet_training_seed_06	.021
146	alexnet_training_seed_01	.020
147	alexnet_training_seed_10	.019
148	alexnet_training_seed_02	.018
149	alexnet_training_seed_09	.016
150	compact_resnet50_robust_V4	.015
151	alexnet_training_seed_07	.014
152	alexnet_training_seed_03	.011
153	mehrer2020_alexnet_training_seed_05
154	mehrer2020_alexnet_training_seed_02
155	mehrer2020_alexnet_training_seed_01_fov12
156	resnet34_ecoset_full
157	ReAlnet01_cornet_fov12

Benchmark bibtex

@inproceedings{zerbe_laion-fmri_2026,
      title = {{LAION}-{fMRI}: A densely sampled 7T-fMRI dataset providing broad coverage of natural image diversity},
      author = {Zerbe, Josefine and Roth, Johannes and Mell, Maggie Mae and Herholz, Peer and Knapen, Tomas and Hebart, Martin N.},
      year = {2026},
      booktitle = {Vision Sciences Society Annual Meeting},
  }

Ceiling

0.76.

Note that scores are relative to this ceiling.

Data: Zerbe2026_fmri.IT-ood

Metric: ridgecv