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("Hebart2023_fmri.V4-ridgecv")
score = benchmark(my_model)

Benchmark API

Code examples

Model scores

Score Legend Min Alignment Max Alignment

Rank	Model	Score
1	custom_model_cv_18_dagger_408	.459
2	dinov2_vitg14_reg_lc	.396
3	dinov2_vitl14_lc_res336	.385
4	dinov2_vitg14_lc	.370
5	cvt_cvt-21-224-in1k_4_LucyV4	.367
6	resnet50_st_wzc_seed7_2_323	.367
7	vit_large_clip336_openai	.365
8	resnet152_imagenet_full	.363
9	cvt_cvt-21-384-in1k_4_LucyV4	.358
10	resnet50_st_wzc_seed3_2_323	.358
11	deit3_base_patch16_224_fb_in22k_ft_in1k_gram_aligned_dinov3	.355
12	deit3_base_patch16_224_fb_in22k_ft_in1k	.355
13	cvt_cvt-w24-384-in22k_finetuned-in1k_4	.352
14	convnext_xxlarge_clip	.351
15	ReAlnet04_cornet	.351
16	convnext_large_in22k_supervised	.347
17	convnext_large:fb_in22k_ft_in1k	.344
18	convnext_xlarge:fb_in22k_ft_in1k	.346
19	vit_large_patch14_clip_336:openai_ft_in12k_in1k	.343
20	eva02_large_in22k	.341
21	vit_relpos_base_patch16_clsgap_224:sw_in1k	.341
22	cvt_cvt-21-384-in22k_finetuned-in1k_4_LucyV4	.342
23	cvt_cvt-13-224-in1k_4_LucyV4	.342
24	ReAlnet01_cornet	.345
25	ReAlnet03_cornet	.344
26	voneresnet-50	.340
27	shufflenet_v2_x1_0	.339
28	cvt_cvt-13-384-in1k_4_LucyV4	.338
29	resnet50-moclr8deg	.337
30	convnext_tiny:in12k_ft_in1k	.331
31	resnet50_ewconplus_wzc	.329
32	AT_efficientnet-b2	.330
33	AdvProp_efficientnet-b2	.330
34	resnet50-VITO-8deg-cc	.327
35	resnet50-vitoimagevidnet8	.327
36	deit_small_imagenet_full_seed-0	.325
37	vit_large_patch14_clip_224:openai_ft_in1k	.324
38	resnet50_ewconplus_wzc_seed6_2_323	.324
39	resnet_50_v2	.321
40	vit_large_patch14_clip_336:laion2b_ft_in1k	.322
41	deepeisnn_ei_imagenet_resnet18	.322
42	dinov2_vitl14_reg_lc_featread	.322
43	dinov2_vitl14_reg_lc	.322
44	focalnet_tiny_lrf_in1k	.320
45	efficientnet_b2_imagenet_full	.319
46	cvt_cvt-13-384-in22k_finetuned-in1k_4_LucyV4	.317
47	AdvProp_efficientnet-b4	.320
48	resnet50_ewconplus_wzc_seed5_2__323	.315
49	alexnet	.315
50	VOneCORnet-S	.316
51	resnet50-SIN	.316
52	vit_base_patch16_clip_224:openai_ft_in12k_in1k	.313
53	resnet50_ewconplus_wzc_seed2_2__323	.312
54	convnext_tiny_sup	.311
55	Res2Net50_26w_4s	.309
56	mobilenet_v2_1_4_224	.307
57	vit_tiny_r_s16_p8_384:augreg_in21k_ft_in1k	.307
58	AdvProp_efficientnet-b8	.306
59	resnext101_32x32d_wsl	.302
60	resnet50-SIN_IN	.299
61	deit3_base_patch16_224_fb_in1k	.296
62	deit3_base_patch16_224_fb_in1k_gram_aligned_dinov3	.296
63	efficientnet_b0	.298
64	convnext_femto_ols:d1_in1k	.295
65	convnext_tiny_imagenet_full_seed-0	.293
66	resnet_SIN_IN_FT_IN	.293
67	resnet50-SIN_IN_IN	.293
68	convnext_large_clip_laion2b_finelayers	.293
69	resnext101_32x16d_wsl	.292
70	resnext101_32x48d_wsl	.292
71	resnet50_imagenet_full	.290
72	vit_base_patch16_clip_224:openai_ft_in1k	.291
73	resnext101_32x8d_wsl	.289
74	resnet50_st_wzc_seed6_2_323	.288
75	vit_large_patch14_clip_224:openai_ft_in12k_in1k	.289
76	yudixie_wbh_object_class_scaling_0415_imagenet	.288
77	deit_base_imagenet_full_seed-0	.284
78	resnet50-sup	.283
79	resnet50_tutorial	.283
80	resnet_50_v1_spiking	.283
81	resnet_50_v1	.283
82	deit_large_imagenet_full_seed-0	.280
83	antialiased-r50	.281
84	convnext_base_imagenet_full_seed-0	.279
85	voneresnet-50-non_stochastic	.278
86	convnext_large_imagenet_full_seed-0	.276
87	vit_relpos_base_patch32_plus_rpn_256:sw_in1k	.275
88	vit_large_patch14_clip_224:laion2b_ft_in1k	.277
89	predictive_coding_pc	.270
90	convnext_small_imagenet_full_seed-0	.270
91	convnext_base:clip_laiona_augreg_ft_in1k_384	.270
92	dinov2_vitl14_lc	.268
93	AlexNet_SIN_fov12	.265
94	swin_small_patch4_window7_224:ms_in22k_ft_in1k	.265
95	pr1_stl10_disentangled_alpha04	.263
96	ReAlnet01_cornet_fov4	.264
97	convnext_large_mlp:clip_laion2b_augreg_ft_in1k_384	.261
98	AlexNet_SIN	.261
99	effnetb1_cutmixpatch_augmix_robust32_avge4e7_manylayers_324x288	.259
100	vit_large_patch14_clip_224:laion2b_ft_in12k_in1k	.260
101	effnetb1_cutmixpatch_SAM_robust32_avge6e8e9e10_manylayers_324x288	.256
102	grcnn	.254
103	resnet-50-robust	.252
104	resnet50_robust_l2_eps1	.249
105	vit_huge_patch14_clip_224:laion2b_ft_in12k_in1k	.247
106	densenet-121	.244
107	convnext_base_fb_in1k_gram_aligned_dinov3	.241
108	convnext_base_fb_in1k	.241
109	AdvProp_efficientnet-b7	.235
110	resnet18_fair	.229
111	AdvProp_efficientnet-b6	.225
112	convnext_base_fb_in22k_ft_in1k_gram_aligned_dinov3	.218
113	convnext_base_fb_in22k_ft_in1k	.218
114	AlexNet_SIN_fov16	.217
115	effnetb1_272x240	.217
116	deepeisnn_ei_cifar10_resnet18	.212
117	deepeisnn_ei_cifar10_vgg19	.207
118	ViT_L_32_imagenet1k	.200
119	yudixie_wbh_object_class_scaling_0415_run_2	.147
120	effnetb1_cutmix_augmix_sam_e1_5avg_424x377	.134
121	alexnet_training_seed_06	.113
122	alexnet_training_seed_02	.111
123	alexnet_training_seed_03	.108
124	yudixie_wbh_object_class_scaling_0415_run_1	.106
125	alexnet_training_seed_09	.102
126	yudixie_wbh_object_class_scaling_0415_run_0	.104
127	alexnet_training_seed_01	.101
128	alexnet_training_seed_07	.098
129	alexnet_training_seed_10	.096
130	hmax	.096
131	compact_resnet50_robust_V4	.077
132	yudixie_wbh_object_class_scaling_0415_run_3	.072
133	vit_huge_clip336_laion
134	convnext_xxlarge:clip_laion2b_soup_ft_in1k
135	pnasnet_large_halfPrecision

Benchmark bibtex

None

Ceiling

0.41.

Note that scores are relative to this ceiling.

Data: Hebart2023_fmri.V4

Metric: ridgecv