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.V2-rdm-pearson")
score = benchmark(my_model)

Benchmark API

Code examples

Model scores

Score Legend Min Alignment Max Alignment

Rank	Model	Score
1	deit_small_imagenet_full_seed-0	.298
2	focalnet_tiny_lrf_in1k	.289
3	resnet50_robust_l2_eps1	.284
4	alexnet	.284
5	vit_large_clip336_openai	.283
6	effnetb1_272x240	.283
7	resnet50-SIN_IN	.282
8	antialiased-r50	.282
9	vit_tiny_r_s16_p8_384:augreg_in21k_ft_in1k	.282
10	AdvProp_efficientnet-b6	.280
11	AdvProp_efficientnet-b4	.278
12	resnet_SIN_IN_FT_IN	.277
13	resnet50-SIN_IN_IN	.277
14	convnext_large:fb_in22k_ft_in1k	.269
15	dinov2_vitl14_lc	.268
16	resnet50-SIN	.266
17	resnet50_ewconplus_wzc	.266
18	voneresnet-50-non_stochastic	.263
19	AdvProp_efficientnet-b8	.261
20	vit_large_patch14_clip_336:laion2b_ft_in1k	.259
21	pnasnet_large_halfPrecision	.259
22	dinov2_vitl14_lc_res336	.257
23	resnet_50_v1_spiking	.257
24	resnet_50_v1	.257
25	resnet50-sup	.257
26	resnet152_imagenet_full	.257
27	voneresnet-50	.256
28	cvt_cvt-21-224-in1k_4_LucyV4	.256
29	Res2Net50_26w_4s	.255
30	convnext_large_imagenet_full_seed-0	.255
31	AlexNet_SIN	.254
32	AdvProp_efficientnet-b7	.252
33	swin_small_patch4_window7_224:ms_in22k_ft_in1k	.252
34	cvt_cvt-w24-384-in22k_finetuned-in1k_4	.250
35	effnetb1_cutmix_augmix_sam_e1_5avg_424x377	.250
36	vit_large_patch14_clip_224:openai_ft_in12k_in1k	.246
37	custom_model_cv_18_dagger_408	.240
38	vit_large_patch14_clip_336:openai_ft_in12k_in1k	.239
39	vit_large_patch14_clip_224:openai_ft_in1k	.236
40	convnext_xlarge:fb_in22k_ft_in1k	.235
41	grcnn	.234
42	deit_large_imagenet_full_seed-0	.234
43	vit_large_patch14_clip_224:laion2b_ft_in1k	.231
44	AlexNet_SIN_fov12	.231
45	dinov2_vitg14_lc	.229
46	convnext_tiny_imagenet_full_seed-0	.230
47	deit_base_imagenet_full_seed-0	.229
48	resnet-50-robust	.228
49	resnet50_imagenet_full	.225
50	convnext_femto_ols:d1_in1k	.225
51	AlexNet_SIN_fov16	.222
52	resnet50-moclr8deg	.221
53	cvt_cvt-13-384-in1k_4_LucyV4	.218
54	resnet50-VITO-8deg-cc	.218
55	resnet50-vitoimagevidnet8	.218
56	vit_base_patch16_clip_224:openai_ft_in12k_in1k	.217
57	vit_large_patch14_clip_224:laion2b_ft_in12k_in1k	.217
58	AdvProp_efficientnet-b2	.215
59	AT_efficientnet-b2	.215
60	cvt_cvt-13-384-in22k_finetuned-in1k_4_LucyV4	.205
61	efficientnet_b0	.203
62	convnext_tiny_sup	.202
63	vit_relpos_base_patch16_clsgap_224:sw_in1k	.197
64	densenet-121	.190
65	convnext_base_imagenet_full_seed-0	.189
66	cvt_cvt-13-224-in1k_4_LucyV4	.188
67	convnext_base:clip_laiona_augreg_ft_in1k_384	.188
68	convnext_small_imagenet_full_seed-0	.187
69	cvt_cvt-21-384-in1k_4_LucyV4	.186
70	vit_relpos_base_patch32_plus_rpn_256:sw_in1k	.187
71	vit_huge_patch14_clip_224:laion2b_ft_in12k_in1k	.186
72	vit_base_patch16_clip_224:openai_ft_in1k	.186
73	resnext101_32x8d_wsl	.179
74	mobilenet_v2_1_4_224	.175
75	efficientnet_b2_imagenet_full	.173
76	convnext_large_mlp:clip_laion2b_augreg_ft_in1k_384	.170
77	resnext101_32x48d_wsl	.167
78	resnet_50_v2	.166
79	convnext_tiny:in12k_ft_in1k	.167
80	resnext101_32x32d_wsl	.164
81	ReAlnet03_cornet	.144
82	cvt_cvt-21-384-in22k_finetuned-in1k_4_LucyV4	.144
83	ReAlnet04_cornet	.142
84	resnet50_tutorial	.142
85	ReAlnet01_cornet	.137
86	alexnet_training_seed_06	.136
87	VOneCORnet-S	.133
88	alexnet_training_seed_02	.125
89	dinov2_vitl14_reg_lc_featread	.113
90	dinov2_vitl14_reg_lc	.113
91	shufflenet_v2_x1_0	.108
92	alexnet_training_seed_01	.109
93	alexnet_training_seed_03	.105
94	compact_resnet50_robust_V4	.103
95	ReAlnet01_cornet_fov4	.102
96	alexnet_training_seed_07	.101
97	alexnet_training_seed_09	.099
98	convnext_xxlarge_clip	.092
99	alexnet_training_seed_10	.090
100	convnext_xxlarge:clip_laion2b_soup_ft_in1k	.086
101	ViT_L_32_imagenet1k	.081
102	effnetb1_cutmixpatch_SAM_robust32_avge6e8e9e10_manylayers_324x288	.067
103	effnetb1_cutmixpatch_augmix_robust32_avge4e7_manylayers_324x288	.066
104	dinov2_vitg14_reg_lc	.064
105	hmax	.053
106	vit_huge_clip336_laion
107	eva02_large_in22k
108	resnext101_32x16d_wsl

Benchmark bibtex

None

Ceiling

0.77.

Note that scores are relative to this ceiling.

Data: Zerbe2026_fmri.V2-rdm

Metric: pearson