Sample stimuli

How to use

from brainscore_vision import load_benchmark
benchmark = load_benchmark("Marques2020_Schiller1976-sf_bandwidth")
score = benchmark(my_model)

Benchmark API

Code examples

Model scores

Score Legend Min Alignment Max Alignment

Rank	Model	Score
1	alexnet_early_checkpoint	.985
2	vonegrcnn_47e	.977
3	yudixie_resnet50_translation_rotation_0_240908	.970
4	resnet101_ecoset_full	.969
5	effnetb1_cutmixrespatch_SAM_robust32_e2_manylayers_324x288	.967
6	effnetb1_cutmixrespatch_SAM_robust32_e10_manylayers_324x288	.965
7	effnetb1_cutmixpatch_augmix_robust32_e5_324x288	.966
8	effnetb1_cutmixpatch_augmix_robust32_avge4e7_manylayers_324x288	.967
9	effnetb1_cutmixpatch_SAM_robust32_avge6e8e9e10_manylayers_324x288	.966
10	0.5x_resnet-18_LC_w_sh_1_iter	.965
11	fulltest_microblockvf_nobottleneck_freeat_nopretrain_eps2_m4	.966
12	yudixie_resnet50_category_class_0_240908	.959
13	resnet152_imagenet_full	.957
14	Res2Net50_26w_4s	.956
15	resnet-18-LC_w_sh_1_iter_m	.953
16	resnet-18-LC_w_sh_100_iter_m	.948
17	resnet101	.945
18	yudixie_resnet18_cat_obj_class_all_latents_0_240719	.946
19	yudixie_resnet18_object_class_0_240719	.945
20	resnet-18-LC_w_sh_1_iter_conv_init_m	.942
21	resnet50_FractalDB	.941
22	vonegrcnn_52e_full	.938
23	resnet34_ecoset_full	.939
24	ReAlnet03_cornet	.938
25	AlexNet_SIN	.938
26	efficientnet_b1_imagenet_full	.938
27	ReAlnet09	.936
28	ReAlnet09_cornet	.939
29	ReAlnet03	.939
30	alexnet	.938
31	alexnet	.934
32	alexnet	.933
33	alexnet	.934
34	alexnet	.934
35	alexnet-baseline	.932
36	alexnet	.934
37	ReAlnet08_cornet	.932
38	alexnet	.930
39	BiT-S-R152x2	.931
40	resnext101_32x8d_wsl	.931
41	ReAlnet08	.931
42	r3m_resnet50_nocrop	.928
43	voneresnet-50-non_stochastic	.929
44	resnet18_imagenet_full	.928
45	vggface	.931
46	ReAlnet10_cornet	.928
47	resnet18_ecoset_full	.926
48	mobilenet_v2_1-4_224_pytorch	.927
49	ReAlnet04_cornet	.928
50	ReAlnet10	.925
51	ReAlnet07_cornet	.923
52	ReAlnet04	.925
53	ReAlnet07	.926
54	ReAlnet05	.924
55	artResNet18_1	.926
56	mobilenet_v2_1.0_96	.921
57	yudixie_resnet18_random_0_240719	.924
58	r3m_resnet50	.924
59	ReAlnet05_cornet	.923
60	nasnet_large-tf	.922
61	resnet50-SIN_IN	.922
62	resnet-18-LC_conv_init_m	.922
63	BiT-M-R50x3	.920
64	alexnet_random_linf8_perturb	.920
65	ReAlnet_incorrect	.920
66	ReAlnet01	.920
67	ReAlnet06	.919
68	ReAlnet06_cornet	.920
69	inception_v2-tf	.919
70	ReAlnet01_cornet	.917
71	ReAlnet02_cornet	.918
72	resnet152_ecoset_full	.916
73	0.5x_resnet-18	.915
74	yudixie_resnet18_rotation_reg_0_240719	.915
75	ReAlnet02	.915
76	antialiased-rnext101_32x8d	.914
77	convnext_tiny:in12k_ft_in1k	.914
78	CORnet-S	.912
79	resnet101_imagenet_full	.911
80	cornet_s	.909
81	efficientnet_b2_imagenet_full	.907
82	BiT-M-R152x2	.907
83	nasnet_mobile	.906
84	mobilenet_v2_1.0_128-tf	.903
85	vit_relpos_base_patch32_plus_rpn_256:sw_in1k	.905
86	yudixie_resnet50_rotation_reg_0_240908	.905
87	mobilenet_v2_0.5_224-tf	.900
88	convnext_xlarge:fb_in22k_ft_in1k	.902
89	BiT-M-R101x3	.900
90	mobilenet_v2_0.5_192-tf	.899
91	r3m_resnet18	.902
92	alexnet_training_seed_01	.901
93	resnet-50_untrained	.901
94	texture_shape_resnet50_trained_on_SIN	.900
95	alexnet_training_seed_10	.900
96	resnet-18-LC_w_sh_100_iter_conv_init_m	.899
97	resnet-50_v1-tf	.898
98	0.5x_resnet-18_LC	.899
99	resnet50-SIN	.898
100	resnet50_finetune_cutmix_e3_robust_linf8255_e0_247x234	.897
101	RN50	.898
102	CLIP_resnet50	.894
103	alexnet_robust_correct	.896
104	resnet50_finetune_cutmix_AVGe2e3_robust_linf8255_e0_247x234	.896
105	CLIP_resnet50_float32	.895
106	nasnet_mobile-tf	.895
107	resnet-18-LC_m	.894
108	r3m_resnet34	.894
109	AdvProp_efficientnet-b0	.890
110	yudixie_resnet18_distance_reg_0_240719	.890
111	grcnn_robust_v1	.888
112	voneresnet-50	.887
113	0.5x_resnet-18_LC_w_sh_100_iter	.890
114	antialiased-r50	.887
115	antialias-resnet152	.887
116	resnet50	.885
117	yudixie_resnet50_translation_reg_0_240908	.886
118	grcnn_v2_text_noise	.885
119	deit_small_imagenet_100_seed-0	.882
120	alexnet_training_seed_03	.883
121	BiT-S-R101x3	.882
122	alexnet_training_seed_07	.882
123	resnet50_ecoset_full	.880
124	yudixie_resnet50_cat_obj_class_all_latents_0_240908	.881
125	alexnet_training_seed_09	.881
126	resnet50_random_linf8_perturb	.877
127	BiT-S-R152x4	.876
128	voneresnet-50-robust	.876
129	mobilenet_v2_1.0_160-tf	.874
130	resnet-152_v2-tf	.874
131	densenet-169-keras	.874
132	resnet34_imagenet_full	.875
133	SWSL_resnext101_32x8d	.873
134	VOneCORnet-S	.870
135	resnet50-vicregl0p9	.871
136	resnet50_byol	.870
137	mobilenet_v2_1.0_192-tf	.868
138	yudixie_resnet18_category_class_0_240719	.868
139	vgg_16	.868
140	yudixie_resnet18_distance_rotation_0_240719	.867
141	focalnet_tiny_lrf_in1k	.864
142	grcnn_109	.864
143	convnext_large:fb_in22k_ft_in1k	.863
144	xception	.861
145	vgg-16-keras	.859
146	yudixie_resnet18_translation_rotation_0_240719	.859
147	resnet50	.858
148	resnet_50_v1	.856
149	resnet50_tutorial	.857
150	yudixie_resnet50_imagenet1kpret_0_240908	.854
151	inception_v3-tf	.855
152	resnet-18_untrained	.854
153	resnet50-sup	.852
154	yudixie_resnet50_distance_translation_0_240908	.855
155	resnet-18-LC_d_w_sh_1x1_conv_init_m	.852
156	resnet50_random_l2_perturb	.849
157	BiT-S-R101x1	.847
158	resnet_50_v2	.846
159	BiT-S-R50x3	.846
160	alexnet_training_seed_02	.847
161	effnetb1_272x240	.847
162	resnet-18-LC_w_sh_10_iter_conv_init_m	.845
163	inception_v1	.845
164	yudixie_resnet50_distance_translation_rotation_0_240908	.845
165	AdvProp_efficientnet-b8	.842
166	alexnet_reduced_aliasing_early_checkpoint	.842
167	resnet50-vicregl0p75	.840
168	resnet50_imagenet_full	.841
169	BiT-S-R50x1	.839
170	yudixie_resnet50_object_class_0_240908	.840
171	vit_large_patch14_clip_224:laion2b_ft_in1k	.840
172	resnet50-SIN_IN_IN	.838
173	yudixie_resnet18_imagenet1kpret_0_240719	.836
174	AdvProp_efficientnet-b6	.836
175	0.5x_resnet-18_LC_w_sh_10_iter	.836
176	resnet-152_v1-tf	.834
177	resnet50_simclr	.835
178	deit_large_imagenet_full_seed-0	.834
179	BiT-M-R50x1	.833
180	resnet50_moco_v2	.834
181	SWSL_resnet50	.830
182	yudixie_resnet50_distance_rotation_0_240908	.831
183	resnet-152_v2_pytorch	.830
184	effnetb1_cutmix_augmix_sam_e1_5avg_424x377	.828
185	cvt_cvt-13-224-in1k_4_LucyV4	.828
186	CLIP-RN50	.828
187	barlow-twins-resnet50	.828
188	densenet-169	.827
189	inception_v3	.825
190	mobilenet_v1_0.25_128	.824
191	alexnet_l2_3_robust	.820
192	resnet-18-LC_1st_conv_m	.818
193	vit_base_patch16_clip_224:openai_ft_in12k_in1k	.814
194	konkle_alexnetgn_ipcl_ref01_primary_model	.813
195	yudixie_resnet18_distance_translation_rotation_0_240719	.814
196	alexnet_training_seed_04	.814
197	deit_base_imagenet_full_seed-0	.813
198	ViT_L_32_imagenet1k	.815
199	vit_large_patch14_clip_224:openai_ft_in12k_in1k	.810
200	BiT-M-R101x1	.809
201	AdvProp_efficientnet-b7	.809
202	resnet50-simclr-vissl	.805
203	grcnn	.804
204	yudixie_resnet50_random_0_240908	.803
205	yudixie_resnet18_distance_translation_0_240719	.803
206	resnet-101_v1-tf	.795
207	alexnet_training_seed_05	.794
208	mobilenet_v2_0.75_96	.793
209	CORnet-Z	.795
210	resnet-50x2_untrained	.794
211	resnet-18-LC_w_sh_10_iter_m	.793
212	yudixie_resnet50_distance_reg_0_240908	.792
213	vit_base_patch16_clip_224:openai_ft_in1k	.790
214	alexnet_training_seed_08	.787
215	konkle_alexnetgn_ipcl_ref12_supervised_ipcl_aug	.789
216	efficientnet_b0	.787
217	resnet50-moclr8deg	.785
218	inception_v1-tf	.785
219	resnet-18_test_m	.784
220	densenet-121	.783
221	resnet50-vicreg	.783
222	cvt_cvt-21-224-in1k_4_LucyV4	.782
223	AT_efficientnet-b2	.782
224	tf_efficientnetv2_s_in21ft1k_robust_linf12255_400x400	.780
225	blt_vs	.780
226	AdvProp_efficientnet-b2	.779
227	vit_large_patch14_clip_224:laion2b_ft_in12k_in1k	.778
228	mobilenet_v2_0.75_192-tf	.776
229	inception_v4-tf	.777
230	mobilenet_v2_1.4_224-tf	.773
231	vit_large_patch14_clip_224:openai_ft_in1k	.773
232	resnet50_imagenet_100_seed-0	.773
233	pnasnet_large-tf	.767
234	ViT-B/32	.764
235	vit_huge_patch14_clip_224:laion2b_ft_in12k_in1k	.766
236	mobilenet_v2_1_4_224	.765
237	resnet-50_v2-tf	.763
238	vision_transformer_vit_large_patch16_224	.762
239	resnet50-barlow	.755
240	resnet50_imagenet_10_seed-0	.749
241	vonegrcnn_62e_nobn	.749
242	omnivore_swinT	.747
243	mobilenet_v2_0.75_128	.742
244	my-model	.743
245	vit_large_patch14_clip_336:laion2b_ft_in1k	.742
246	yudixie_resnet18_translation_reg_0_240719	.739
247	v1-pyr-nodown	.735
248	deit_small_imagenet_full_seed-0	.736
249	AdvProp_efficientnet-b4	.728
250	shufflenet_v2_x1_0	.729
251	resnet50_linf_4_robust	.727
252	alexnet_linf_8_robust	.724
253	densenet-121-keras	.724
254	efficientnet-b7	.723
255	vit_relpos_base_patch16_clsgap_224:sw_in1k	.724
256	texture_shape_alexnet_trained_on_SIN	.723
257	resnet_SIN_IN_FT_IN	.722
258	vgg_19	.718
259	vonealexnet_gaussian_noise_std4_fixed	.717
260	mobilenet_v2_1_0_128	.718
261	cv_18_dagger_408_pretrained	.716
262	mobilenet_v2_1.0_224-tf	.714
263	resnet50-VITO-8deg-cc	.714
264	vit_huge_patch14_clip_336:laion2b_ft_in12k_in1k	.712
265	convnext_large_imagenet_full_seed-0	.712
266	inception_resnet_v2	.708
267	cvt_cvt-13-384-in22k_finetuned-in1k_4_LucyV4	.704
268	BiT-M-R152x4	.694
269	effnetv2m_custom384	.690
270	custom_model_cv_18_dagger_408	.685
271	mobilenet_v2_0_5_224	.680
272	resnet-101_v2-tf	.674
273	resnet50-vitoimagevidnet8	.674
274	CORnetZ_CIFAR10_bs32_20_04	.670
275	cvt_cvt-21-384-in1k_4_LucyV4	.666
276	resnet50_l2_3_robust	.667
277	resnet50_robust_l2_eps1	.665
278	convnext_base:clip_laiona_augreg_ft_in1k_384	.664
279	resnet18_imagenet21kP	.664
280	convnext_xxlarge:clip_laion2b_soup_ft_in1k	.663
281	resnet50-cifar	.661
282	mobilenet_v2_0.5_128	.660
283	hmax	.662
284	convnext_tiny_imagenet_full_seed-0	.661
285	mobilenet_v2_0_75_224	.655
286	resnet50_robust_l2_eps3	.653
287	mobilenet_v2_0_5_192	.650
288	convnext_small_imagenet_100_seed-0	.649
289	CLIP_ViT-B_32	.646
290	cvt_cvt-13-384-in1k_4_LucyV4	.644
291	imagenet_l2_3_0	.643
292	imagenet_l2_10_0	.640
293	alexnet	.632
294	mobilenet_v2_0.75_224-tf	.626
295	vgg-19-keras	.617
296	mobilenet_v2_0_75_192	.615
297	omnivore_swinB	.609
298	vit_large_patch14_clip_336:openai_ft_in12k_in1k	.607
299	densenet-201	.604
300	cvt_cvt-w24-384-in22k_finetuned-in1k_4	.601
301	resnet50_linf_8_robust	.596
302	resnet-50-robust	.596
303	mobilenet_v2_1_0_160	.595
304	densenet-201-keras	.590
305	swin_small_patch4_window7_224:ms_in22k_ft_in1k	.586
306	omnivore_swinS	.585
307	vit_tiny_r_s16_p8_384:augreg_in21k_ft_in1k	.584
308	mobilenet_v2_0.35_224	.572
309	cvt_cvt-21-384-in22k_finetuned-in1k_4_LucyV4	.570
310	FrankRobWobv0	.567
311	convnext_femto_ols:d1_in1k	.557
312	alexnet_random_l2_3_perturb	.552
313	mobilenet_v2_1_0_224	.550
314	convnext_base_imagenet_full_seed-0	.546
315	convnext_large_mlp:clip_laion2b_augreg_ft_in1k_384	.532
316	mobilenet_v2_1_0_192	.523
317	mobilenet_v2_0_75_160	.520
318	convnext_small_imagenet_full_seed-0	.501
319	CORnetZ_CIFAR10	.495
320	efficientnet_b0_imagenet_full	.491
321	convnext_small_imagenet_10_seed-0	.488
322	resnet50_imagenet_1_seed-0	.481
323	mobilevit_small	.477
324	mobilenet_v2_0.75_160-tf	.471
325	resnet50-meshes-lt-100-original-pretrained	.456
326	resnet-18-LC_untrained	.454
327	mobilenet_v2_1_3_224	.442
328	alexnet_training_seed_06	.428
329	mobilenet_v2_1.3_224-tf	.399
330	resnet50_pretrained_with_retinal_waves	.397
331	resnet50-meshes-lt-100-original-scratch	.338
332	convnext_tiny_sup	.331
333	convnext_small_imagenet_1_seed-0	.293
334	my-model	.226
335	mobilenet_v2_0.5_160	.103
336	deit_small_imagenet_10_seed-0	.000
337	regnet_y_400mf	X
338	nasnet_large	X
339	mobilenet_v2_0.35_160	X
340	resnet_101_v2	X
341	deit_small_imagenet_1_seed-0	X
342	resnet_152_v1	X
343	openclip	X
344	vonegrcnn_62e_nobn	X
345	vonegrcnn_47e	X
346	vonegrcnn_52e_full	X
347	inception_v4	X
348	tv_efficientnet-b1	X
349	alexnet_ks_torevert	X
350	fitvid_trained_on_physion	X
351	resnet18-supervised	X
352	resnet50_ImageNet	X
353	yudixie_resnet50_imagenet1kpret_0_240312	X
354	pnasnet_large	X
355	densenet_201_pytorch	X
356	resnext101_32x48d_wsl	X
357	resnext101_32x16d_wsl	X
358	resnext101_32x32d_wsl	X
359	pnasnet_large_pytorch	X
360	inception_v3_pytorch	X
361	resnet_152_v2	X
362	resnet_101_v1	X

Benchmark bibtex

@article{Schiller1976,
            author = {Schiller, P. H. and Finlay, B. L. and Volman, S. F.},
            doi = {10.1152/jn.1976.39.6.1352},
            issn = {0022-3077},
            journal = {Journal of neurophysiology},
            number = {6},
            pages = {1334--1351},
            pmid = {825624},
            title = {{Quantitative studies of single-cell properties in monkey striate cortex. III. Spatial Frequency}},
            url = {http://www.ncbi.nlm.nih.gov/pubmed/825624},
            volume = {39},
            year = {1976}
            }

Ceiling

0.93.

Note that scores are relative to this ceiling.

Data: Marques2020_Schiller1976

Metric: sf_bandwidth