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.V1-ood-ridgecv")
score = benchmark(my_model)

Benchmark API

Code examples

Model scores

Score Legend Min Alignment Max Alignment

Rank	Model	Score
1	vit_huge_clip336_laion	.425
2	cvt_cvt-13-384-in1k_4_LucyV4	.424
3	cvt_cvt-21-384-in22k_finetuned-in1k_4_LucyV4	.415
4	cvt_cvt-w24-384-in22k_finetuned-in1k_4	.414
5	vit_relpos_base_patch16_clsgap_224:sw_in1k	.407
6	vit_tiny_r_s16_p8_384:augreg_in21k_ft_in1k	.404
7	densenet-121	.402
8	resnext101_32x16d_wsl	.400
9	convnext_large:fb_in22k_ft_in1k	.397
10	shufflenet_v2_x1_0	.397
11	vit_large_clip336_openai	.397
12	convnext_xxlarge:clip_laion2b_soup_ft_in1k	.393
13	vit_large_patch14_clip_224:openai_ft_in1k	.394
14	cvt_cvt-13-384-in22k_finetuned-in1k_4_LucyV4	.392
15	AdvProp_efficientnet-b6	.391
16	resnet_50_v1_spiking	.391
17	resnet50-sup	.391
18	resnet_50_v1	.391
19	resnet50_tutorial	.391
20	resnext101_32x48d_wsl	.390
21	resnet50-SIN_IN_IN	.390
22	vit_large_patch14_clip_224:openai_ft_in12k_in1k	.388
23	regnety_032	.368
24	convnext_femto_ols:d1_in1k	.386
25	antialiased-r50	.384
26	grcnn	.382
27	convnext_xlarge:fb_in22k_ft_in1k	.380
28	focalnet_tiny_lrf_in1k	.381
29	efficientnet_b1_imagenet_full	.362
30	cvt_cvt-13-224-in1k_4_LucyV4	.379
31	efficientnet_b2_imagenet_full	.377
32	cvt_cvt-21-224-in1k_4_LucyV4	.377
33	AdvProp_efficientnet-b4	.375
34	cvt_cvt-21-384-in1k_4_LucyV4	.376
35	resnet50-SIN	.376
36	custom_model_cv_18_dagger_408	.375
37	resnet_SIN_IN_FT_IN	.376
38	effnetb1_272x240	.374
39	seresnext50_32x4d	.355
40	convnext_tiny_sup	.371
41	effnetb1_cutmix_augmix_sam_e1_5avg_424x377	.370
42	convnext_large_mlp:clip_laion2b_augreg_ft_in1k_384	.370
43	pnasnet_large_halfPrecision	.369
44	vit_huge_patch14_clip_224:laion2b_ft_in12k_in1k	.371
45	convnext_base_imagenet_full_seed-0	.365
46	efficientnet_b0_imagenet_full	.350
47	convnext_small_imagenet_full_seed-0	.362
48	dinov2_vitg14_reg_lc	.363
49	resnet50_simsiam_linhead	.345
50	mobilenet_v2_1_4_224	.361
51	gabor_filter_entropy_4x4_v2_forcedCosine	.345
52	resnet50-moclr8deg	.359
53	AdvProp_efficientnet-b7	.360
54	vit_large_patch14_clip_224:laion2b_ft_in12k_in1k	.360
55	gabor_filter_entropy_4x4_v2	.342
56	vit_large_patch14_clip_336:openai_ft_in12k_in1k	.357
57	dinov2_vitg14_lc	.357
58	voneresnet-50	.356
59	vit_large_patch14_clip_336:laion2b_ft_in1k	.353
60	vit_base_patch16_clip_224:openai_ft_in12k_in1k	.352
61	vit_large_patch14_clip_224:laion2b_ft_in1k	.354
62	resnet50_robust_l2_eps1	.350
63	efficientnet_b0	.350
64	dinov2_vitl14_reg_lc	.350
65	eva02_large_in22k	.349
66	AT_efficientnet-b2	.347
67	AdvProp_efficientnet-b2	.347
68	vit_base_patch16_clip_224:openai_ft_in1k	.348
69	dinov2_vitl14_lc	.349
70	resnext101_32x32d_wsl	.346
71	ijepa_vith14_linhead	.333
72	convnext_tiny:in12k_ft_in1k	.344
73	deit_base_imagenet_full_seed-0	.334
74	resnet50_byol_linhead	.327
75	swin_small_patch4_window7_224:ms_in22k_ft_in1k	.343
76	resnet50_barlowtwins_linhead	.326
77	resnext101_32x8d_wsl	.340
78	resnet50-VITO-8deg-cc	.340
79	resnet50-vitoimagevidnet8	.340
80	AdvProp_efficientnet-b8	.338
81	vit_relpos_base_patch32_plus_rpn_256:sw_in1k	.341
82	ReAlnet01_cornet_fov16	.325
83	dinov2_vitl14_lc_res336	.336
84	resnet_50_v2	.334
85	convnext_tiny_imagenet_full_seed-0	.332
86	convnext_large_imagenet_full_seed-0	.331
87	resnet-50-robust	.333
88	deit_small_imagenet_full_seed-0	.320
89	convnext_base:clip_laiona_augreg_ft_in1k_384	.330
90	ReAlnet01_cornet_fov12	.314
91	mehrer2020_alexnet_training_seed_01	.310
92	gabor_filter_entropy_4x4	.307
93	resnet50-SIN_IN	.321
94	deit_small_imagenet_100_seed-0	.308
95	deit_large_imagenet_full_seed-0	.308
96	mehrer2020_alexnet_training_seed_08	.305
97	mehrer2020_alexnet_training_seed_09	.305
98	ReAlnet10_cornet	.305
99	ReAlnet08_cornet	.304
100	Res2Net50_26w_4s	.317
101	mehrer2020_alexnet_training_seed_10	.302
102	ReAlnet02_cornet	.303
103	ReAlnet07_cornet	.303
104	mehrer2020_alexnet_training_seed_06	.300
105	ReAlnet06_cornet	.301
106	ReAlnet05_cornet	.300
107	ReAlnet03_cornet	.312
108	mehrer2020_alexnet_training_seed_01_fov12	.299
109	ReAlnet01_cornet	.312
110	ReAlnet09_cornet	.299
111	ReAlnet04_cornet	.312
112	mehrer2020_alexnet_training_seed_07	.298
113	effnetb1_cutmixpatch_SAM_robust32_avge6e8e9e10_manylayers_324x288	.310
114	effnetb1_cutmixpatch_augmix_robust32_avge4e7_manylayers_324x288	.310
115	resnet50_ewconplus_wzc	.308
116	mehrer2020_alexnet_training_seed_01_fov16	.289
117	alexnet	.299
118	AlexNet_SIN_fov16	.297
119	ViT_L_32_imagenet1k	.296
120	resnet34_imagenet_full	.281
121	AlexNet_SIN	.288
122	resnet152_imagenet_full	.288
123	AlexNet_SIN_fov12	.284
124	resnet18_imagenet_full	.271
125	resnet101_imagenet_full	.272
126	resnet50_ecoset_full	.268
127	resnet101_ecoset_full	.268
128	resnet152_ecoset_full	.265
129	compact_resnet50_robust_V4	.266
130	eicircuit_resnet18_bn_20260919	.242
131	resnet50_imagenet_100_seed-0	.241
132	resnet18_ecoset_full	.234
133	resnet50_imagenet_full	.236
134	resnet50_simclr_linhead_01norm	.228
135	gabor_filter_mean_abs_4x4_v2	.211
136	alexnet_training_seed_02	.213
137	alexnet_training_seed_06	.206
138	resnet50_vicreg_linhead	.205
139	alexnet_training_seed_09	.203
140	alexnet_training_seed_07	.202
141	alexnet_training_seed_01	.200
142	alexnet_training_seed_10	.196
143	mehrer2020_alexnet_training_seed_01_fov4	.180
144	alexnet_training_seed_03	.178
145	AlexNet_SIN_fov4	.164
146	hmax	.163
147	ReAlnet01_cornet_fov4	.163
148	voneresnet-50-non_stochastic	.130
149	VOneCORnet-S	.084
150	mehrer2020_alexnet_training_seed_05
151	mehrer2020_alexnet_training_seed_04
152	mehrer2020_alexnet_training_seed_03
153	mehrer2020_alexnet_training_seed_02
154	resnet34_ecoset_full
155	convnext_small_imagenet_100_seed-0
156	convnext_xxlarge_clip
157	dinov2_vitl14_reg_lc_featread

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

0.81.

Note that scores are relative to this ceiling.

Data: Zerbe2026_fmri.V1-ood

Metric: ridgecv