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

Benchmark API

Code examples

Model scores

Score Legend Min Alignment Max Alignment

Rank	Model	Score
1	convnext_tiny:in12k_ft_in1k	.463
2	convnext_large:fb_in22k_ft_in1k	.453
3	vit_base_patch16_clip_224:openai_ft_in12k_in1k	.438
4	effnetb1_cutmix_augmix_sam_e1_5avg_424x377	.436
5	convnext_femto_ols:d1_in1k	.433
6	deit_large_imagenet_full_seed-0	.429
7	deit_base_imagenet_full_seed-0	.425
8	resnet50-moclr8deg	.421
9	vit_large_patch14_clip_224:laion2b_ft_in12k_in1k	.414
10	convnext_base_imagenet_full_seed-0	.412
11	shufflenet_v2_x1_0	.409
12	vit_large_patch14_clip_224:openai_ft_in1k	.406
13	convnext_large_imagenet_full_seed-0	.402
14	resnet50_imagenet_full	.398
15	vit_relpos_base_patch32_plus_rpn_256:sw_in1k	.398
16	VOneCORnet-S	.397
17	vit_large_patch14_clip_336:openai_ft_in12k_in1k	.395
18	ReAlnet04_cornet	.391
19	custom_model_cv_18_dagger_408	.390
20	ReAlnet03_cornet	.389
21	ReAlnet01_cornet	.389
22	densenet-121	.389
23	voneresnet-50-non_stochastic	.385
24	effnetb1_272x240	.381
25	resnet50_ewconplus_wzc	.379
26	convnext_xxlarge:clip_laion2b_soup_ft_in1k	.379
27	deit_small_imagenet_full_seed-0	.377
28	alexnet	.375
29	resnet50-SIN_IN	.374
30	convnext_base:clip_laiona_augreg_ft_in1k_384	.371
31	resnet50-SIN_IN_IN	.370
32	resnet50-VITO-8deg-cc	.368
33	resnet50-vitoimagevidnet8	.368
34	resnet50_robust_l2_eps1	.367
35	antialiased-r50	.366
36	voneresnet-50	.362
37	resnet_50_v1_spiking	.361
38	resnet_50_v1	.361
39	resnet50-sup	.361
40	ReAlnet01_cornet_fov4	.357
41	vit_huge_patch14_clip_224:laion2b_ft_in12k_in1k	.355
42	efficientnet_b0	.350
43	resnet152_imagenet_full	.348
44	resnet_SIN_IN_FT_IN	.346
45	dinov2_vitl14_lc	.345
46	resnet_50_v2	.344
47	convnext_small_imagenet_full_seed-0	.345
48	vit_base_patch16_clip_224:openai_ft_in1k	.344
49	convnext_tiny_imagenet_full_seed-0	.344
50	vit_large_patch14_clip_224:laion2b_ft_in1k	.340
51	vit_large_patch14_clip_224:openai_ft_in12k_in1k	.339
52	Res2Net50_26w_4s	.336
53	vit_tiny_r_s16_p8_384:augreg_in21k_ft_in1k	.334
54	mobilenet_v2_1_4_224	.322
55	pnasnet_large_halfPrecision	.320
56	vit_relpos_base_patch16_clsgap_224:sw_in1k	.318
57	convnext_large_mlp:clip_laion2b_augreg_ft_in1k_384	.315
58	AlexNet_SIN	.304
59	cvt_cvt-13-224-in1k_4_LucyV4	.301
60	AdvProp_efficientnet-b2	.301
61	AT_efficientnet-b2	.301
62	vit_large_patch14_clip_336:laion2b_ft_in1k	.300
63	resnet50-SIN	.297
64	effnetb1_cutmixpatch_SAM_robust32_avge6e8e9e10_manylayers_324x288	.297
65	resnet50_tutorial	.294
66	convnext_xlarge:fb_in22k_ft_in1k	.277
67	AdvProp_efficientnet-b7	.274
68	AdvProp_efficientnet-b8	.274
69	cvt_cvt-21-224-in1k_4_LucyV4	.270
70	AdvProp_efficientnet-b4	.268
71	swin_small_patch4_window7_224:ms_in22k_ft_in1k	.267
72	convnext_tiny_sup	.265
73	AdvProp_efficientnet-b6	.261
74	cvt_cvt-21-384-in1k_4_LucyV4	.260
75	cvt_cvt-13-384-in22k_finetuned-in1k_4_LucyV4	.254
76	resnet-50-robust	.245
77	grcnn	.236
78	cvt_cvt-13-384-in1k_4_LucyV4	.222
79	vit_large_clip336_openai	.216
80	dinov2_vitg14_lc	.208
81	dinov2_vitl14_reg_lc_featread	.201
82	dinov2_vitl14_reg_lc	.201
83	ViT_L_32_imagenet1k	.199
84	efficientnet_b2_imagenet_full	.193
85	cvt_cvt-21-384-in22k_finetuned-in1k_4_LucyV4	.184
86	resnext101_32x16d_wsl	.174
87	focalnet_tiny_lrf_in1k	.168
88	convnext_xxlarge_clip	.157
89	dinov2_vitg14_reg_lc	.156
90	resnext101_32x48d_wsl	.154
91	resnext101_32x8d_wsl	.139
92	AlexNet_SIN_fov12	.136
93	resnext101_32x32d_wsl	.129
94	effnetb1_cutmixpatch_augmix_robust32_avge4e7_manylayers_324x288	.118
95	AlexNet_SIN_fov16	.097
96	hmax	.085
97	alexnet_training_seed_06	.049
98	alexnet_training_seed_01	.042
99	alexnet_training_seed_02	.041
100	alexnet_training_seed_03	.041
101	alexnet_training_seed_07	.039
102	compact_resnet50_robust_V4	.038
103	alexnet_training_seed_09	.038
104	alexnet_training_seed_10	.035
105	vit_huge_clip336_laion
106	dinov2_vitl14_lc_res336
107	eva02_large_in22k
108	cvt_cvt-w24-384-in22k_finetuned-in1k_4

Benchmark bibtex

None

Ceiling

0.75.

Note that scores are relative to this ceiling.

Data: Zerbe2026_fmri.IT-rdm

Metric: pearson