Sample stimuli

How to use

from brainscore_vision import load_benchmark
benchmark = load_benchmark("Marques2020_Ringach2002-orth_pref_ratio")
score = benchmark(my_model)

Benchmark API

Code examples

Model scores

Score Legend Min Alignment Max Alignment

Rank	Model	Score
1	resnet-18-LC_m	.983
2	yudixie_resnet50_rotation_reg_0_240908	.974
3	vit_tiny_r_s16_p8_384:augreg_in21k_ft_in1k	.973
4	yudixie_resnet18_category_class_0_240719	.971
5	vonegrcnn_62e_nobn	.958
6	cvt_cvt-21-384-in1k_4_LucyV4	.954
7	0.5x_resnet-18_LC_w_sh_10_iter	.954
8	yudixie_resnet50_distance_translation_rotation_0_240908	.953
9	efficientnet_b1_imagenet_full	.949
10	yudixie_resnet50_cat_obj_class_all_latents_0_240908	.949
11	yudixie_resnet18_cat_obj_class_all_latents_0_240719	.943
12	convnext_large:fb_in22k_ft_in1k	.944
13	convnext_femto_ols:d1_in1k	.941
14	yudixie_resnet50_category_class_0_240908	.938
15	yudixie_resnet18_translation_reg_0_240719	.938
16	effnetb1_cutmixpatch_SAM_robust32_avge6e8e9e10_manylayers_324x288	.937
17	effnetb1_cutmixpatch_augmix_robust32_e5_324x288	.937
18	xception	.937
19	effnetb1_cutmixrespatch_SAM_robust32_e2_manylayers_324x288	.937
20	effnetb1_cutmixpatch_augmix_robust32_avge4e7_manylayers_324x288	.936
21	effnetb1_cutmixrespatch_SAM_robust32_e10_manylayers_324x288	.936
22	cvt_cvt-21-224-in1k_4_LucyV4	.935
23	yudixie_resnet50_distance_rotation_0_240908	.932
24	cvt_cvt-13-224-in1k_4_LucyV4	.933
25	blt_vs	.931
26	yudixie_resnet50_translation_rotation_0_240908	.928
27	mobilenet_v2_1.3_224-tf	.930
28	mobilenet_v2_1_0_128	.929
29	inception_v1	.926
30	resnet-18-LC_w_sh_1_iter_m	.926
31	resnet-18-LC_1st_conv_m	.926
32	resnet-152_v2-tf	.922
33	vgg_19	.921
34	AdvProp_efficientnet-b6	.921
35	alexnet	.919
36	efficientnet-b7	.918
37	AlexNet_SIN	.919
38	BiT-M-R50x3	.916
39	resnet-18-LC_d_w_sh_1x1_conv_init_m	.916
40	fulltest_microblockvf_nobottleneck_freeat_nopretrain_eps2_m4	.915
41	alexnet-baseline	.912
42	alexnet	.912
43	alexnet	.911
44	alexnet	.911
45	yudixie_resnet18_translation_rotation_0_240719	.911
46	alexnet	.911
47	alexnet	.911
48	convnext_base:clip_laiona_augreg_ft_in1k_384	.911
49	alexnet	.910
50	alexnet	.910
51	grcnn	.911
52	alexnet_robust_correct	.909
53	texture_shape_alexnet_trained_on_SIN	.906
54	resnext101_32x8d_wsl	.905
55	alexnet_random_l2_3_perturb	.903
56	cvt_cvt-13-384-in22k_finetuned-in1k_4_LucyV4	.903
57	resnet-18-LC_w_sh_10_iter_conv_init_m	.903
58	focalnet_tiny_lrf_in1k	.901
59	yudixie_resnet18_object_class_0_240719	.901
60	mobilenet_v2_1.0_160-tf	.901
61	yudixie_resnet50_distance_reg_0_240908	.901
62	AdvProp_efficientnet-b8	.900
63	omnivore_swinS	.898
64	Res2Net50_26w_4s	.896
65	BiT-S-R50x1	.895
66	resnet50_imagenet_100_seed-0	.894
67	mobilenet_v2_1.0_224-tf	.893
68	yudixie_resnet50_distance_translation_0_240908	.892
69	resnet-18-LC_w_sh_100_iter_conv_init_m	.893
70	grcnn_robust_v1	.887
71	yudixie_resnet18_distance_rotation_0_240719	.886
72	resnet101	.887
73	konkle_alexnetgn_ipcl_ref12_supervised_ipcl_aug	.885
74	0.5x_resnet-18_LC_w_sh_100_iter	.884
75	resnet-18-LC_w_sh_1_iter_conv_init_m	.879
76	resnet50-vitoimagevidnet8	.876
77	yudixie_resnet50_translation_reg_0_240908	.876
78	omnivore_swinB	.874
79	vgg-16-keras	.874
80	alexnet_reduced_aliasing_early_checkpoint	.874
81	barlow-twins-resnet50	.873
82	resnet18_ecoset_full	.872
83	vgg_16	.871
84	grcnn_109	.870
85	resnet50_byol	.868
86	resnet50-SIN_IN	.867
87	convnext_small_imagenet_full_seed-0	.867
88	CORnet-Z	.867
89	grcnn_v2_text_noise	.866
90	alexnet_random_linf8_perturb	.864
91	resnet_50_v2	.862
92	resnet-50_v1-tf	.861
93	alexnet_l2_3_robust	.860
94	resnet-18-LC_w_sh_10_iter_m	.859
95	0.5x_resnet-18_LC_w_sh_1_iter	.858
96	ReAlnet04	.857
97	swin_small_patch4_window7_224:ms_in22k_ft_in1k	.857
98	ReAlnet04_cornet	.857
99	ReAlnet03	.855
100	ReAlnet03_cornet	.856
101	0.5x_resnet-18_LC	.853
102	BiT-M-R101x1	.853
103	tf_efficientnetv2_s_in21ft1k_robust_linf12255_400x400	.852
104	ReAlnet_incorrect	.850
105	inception_v3-tf	.850
106	CORnetZ_CIFAR10_bs32_20_04	.850
107	ReAlnet01	.850
108	resnet50-vicregl0p9	.850
109	ReAlnet08_cornet	.849
110	ReAlnet08	.850
111	ReAlnet01_cornet	.847
112	ReAlnet09	.846
113	ReAlnet09_cornet	.847
114	convnext_tiny_imagenet_full_seed-0	.847
115	resnet50-simclr-vissl	.847
116	ReAlnet02	.847
117	ReAlnet07	.847
118	ReAlnet02_cornet	.846
119	yudixie_resnet50_object_class_0_240908	.846
120	ReAlnet07_cornet	.846
121	inception_v3	.845
122	alexnet_linf_8_robust	.844
123	yudixie_resnet18_distance_translation_rotation_0_240719	.843
124	BiT-S-R50x3	.844
125	ReAlnet10	.843
126	nasnet_large-tf	.842
127	resnet18_imagenet21kP	.841
128	vonealexnet_gaussian_noise_std4_fixed	.839
129	ReAlnet10_cornet	.839
130	alexnet_early_checkpoint	.838
131	BiT-S-R101x3	.837
132	resnet50_imagenet_10_seed-0	.837
133	ReAlnet06_cornet	.835
134	ReAlnet06	.836
135	ReAlnet05	.833
136	ReAlnet05_cornet	.832
137	resnet101_imagenet_full	.829
138	convnext_tiny_sup	.829
139	antialias-resnet152	.827
140	yudixie_resnet18_distance_translation_0_240719	.826
141	densenet-201	.826
142	convnext_small_imagenet_100_seed-0	.825
143	resnet_50_v1	.824
144	yudixie_resnet50_imagenet1kpret_0_240908	.821
145	resnet50_tutorial	.822
146	resnet50-sup	.821
147	densenet-201-keras	.821
148	AdvProp_efficientnet-b7	.819
149	CORnet-S	.816
150	mobilenet_v2_0_75_192	.817
151	convnext_xlarge:fb_in22k_ft_in1k	.817
152	resnet50-moclr8deg	.814
153	resnet-50_untrained	.815
154	resnet50-vicregl0p75	.812
155	CLIP-RN50	.811
156	imagenet_l2_10_0	.810
157	vgg-19-keras	.810
158	imagenet_l2_3_0	.810
159	RN50	.807
160	resnet50	.807
161	resnet-50x2_untrained	.806
162	mobilenet_v2_1_4_224	.806
163	effnetb1_cutmix_augmix_sam_e1_5avg_424x377	.805
164	inception_v4-tf	.804
165	cornet_s	.803
166	resnet50-SIN_IN_IN	.800
167	resnet-50-robust	.800
168	mobilenet_v2_1_3_224	.799
169	resnet50_robust_l2_eps1	.796
170	resnet50_random_linf8_perturb	.795
171	deit_small_imagenet_full_seed-0	.795
172	r3m_resnet34_nocrop	.795
173	resnet50_l2_3_robust	.792
174	resnet50_linf_8_robust	.787
175	BiT-S-R152x2	.786
176	resnet50_finetune_cutmix_AVGe2e3_robust_linf8255_e0_247x234	.785
177	resnet50_finetune_cutmix_e3_robust_linf8255_e0_247x234	.784
178	resnet50_robust_l2_eps3	.783
179	antialiased-r50	.782
180	resnet50_random_l2_perturb	.780
181	resnet_SIN_IN_FT_IN	.779
182	resnet50-vicreg	.779
183	yudixie_resnet18_imagenet1kpret_0_240719	.779
184	resnet34_imagenet_full	.778
185	resnet18_imagenet_full	.776
186	efficientnet_b0	.775
187	convnext_large_imagenet_full_seed-0	.775
188	convnext_base_imagenet_full_seed-0	.772
189	mobilenet_v2_1.0_96	.772
190	VOneCORnet-S	.772
191	convnext_large_mlp:clip_laion2b_augreg_ft_in1k_384	.769
192	texture_shape_resnet50_trained_on_SIN	.769
193	vit_base_patch16_clip_224:openai_ft_in1k	.769
194	densenet-121	.767
195	mobilenet_v2_0_5_224	.767
196	inception_v2-tf	.766
197	voneresnet-50-robust	.764
198	voneresnet-50	.758
199	mobilenet_v2_0.75_192-tf	.757
200	resnet50_linf_4_robust	.756
201	CLIP_resnet50	.756
202	convnext_tiny:in12k_ft_in1k	.755
203	CLIP_resnet50_float32	.755
204	AdvProp_efficientnet-b4	.754
205	mobilenet_v2_1.4_224-tf	.755
206	densenet-169	.754
207	resnet50_ecoset_full	.752
208	resnet-18_untrained	.749
209	resnet50-VITO-8deg-cc	.747
210	0.5x_resnet-18	.745
211	resnet50-cifar	.743
212	resnet50-SIN	.741
213	SWSL_resnet50	.741
214	resnet50_imagenet_full	.741
215	mobilenet_v2_0_75_224	.739
216	AdvProp_efficientnet-b0	.738
217	resnet152_imagenet_full	.738
218	resnet101_ecoset_full	.735
219	alexnet_training_seed_02	.735
220	resnet-101_v2-tf	.732
221	mobilenet_v2_1.0_192-tf	.730
222	mobilenet_v2_1-4_224_pytorch	.728
223	pnasnet_large-tf	.727
224	alexnet_training_seed_07	.727
225	cvt_cvt-13-384-in1k_4_LucyV4	.725
226	vggface	.725
227	resnet-50_v2-tf	.724
228	resnet-18_test_m	.720
229	vit_base_patch16_clip_224:openai_ft_in12k_in1k	.719
230	alexnet_training_seed_04	.719
231	vonegrcnn_47e	.718
232	resnet-152_v1-tf	.716
233	yudixie_resnet18_distance_reg_0_240719	.715
234	alexnet_training_seed_01	.714
235	mobilenet_v2_0.5_192-tf	.715
236	resnet-18-LC_w_sh_100_iter_m	.715
237	voneresnet-50-non_stochastic	.712
238	mobilenet_v2_0.75_128	.713
239	yudixie_resnet50_random_0_240908	.711
240	alexnet_training_seed_09	.710
241	mobilenet_v1_0.25_128	.707
242	vit_relpos_base_patch32_plus_rpn_256:sw_in1k	.707
243	FrankRobWobv0	.705
244	antialiased-rnext101_32x8d	.705
245	resnet-101_v1-tf	.704
246	BiT-S-R152x4	.702
247	vit_large_patch14_clip_224:laion2b_ft_in1k	.701
248	cvt_cvt-w24-384-in22k_finetuned-in1k_4	.701
249	resnet34_ecoset_full	.700
250	BiT-M-R101x3	.698
251	resnet152_ecoset_full	.697
252	BiT-M-R152x4	.697
253	mobilenet_v2_1_0_192	.696
254	SWSL_resnext101_32x8d	.696
255	convnext_small_imagenet_10_seed-0	.695
256	vonegrcnn_52e_full	.693
257	vit_huge_patch14_clip_336:laion2b_ft_in12k_in1k	.693
258	resnet-152_v2_pytorch	.693
259	densenet-121-keras	.692
260	v1-pyr-nodown	.689
261	efficientnet_b0_imagenet_full	.687
262	deit_small_imagenet_100_seed-0	.686
263	yudixie_resnet18_rotation_reg_0_240719	.687
264	BiT-S-R101x1	.678
265	konkle_alexnetgn_ipcl_ref01_primary_model	.676
266	my-model	.677
267	mobilenet_v2_0.75_160-tf	.669
268	convnext_xxlarge:clip_laion2b_soup_ft_in1k	.668
269	nasnet_mobile-tf	.664
270	resnet50-barlow	.662
271	shufflenet_v2_x1_0	.658
272	BiT-M-R152x2	.658
273	nasnet_mobile	.657
274	vision_transformer_vit_large_patch16_224	.656
275	resnet50_FractalDB	.655
276	CLIP_ViT-B_32	.653
277	densenet-169-keras	.652
278	BiT-M-R50x1	.648
279	effnetv2m_custom384	.640
280	vit_large_patch14_clip_336:openai_ft_in12k_in1k	.639
281	vit_large_patch14_clip_336:laion2b_ft_in1k	.637
282	alexnet_training_seed_10	.636
283	vit_large_patch14_clip_224:laion2b_ft_in12k_in1k	.634
284	cv_18_dagger_408_pretrained	.632
285	resnet50	.630
286	cvt_cvt-21-384-in22k_finetuned-in1k_4_LucyV4	.630
287	mobilenet_v2_0_75_160	.627
288	alexnet_training_seed_03	.627
289	resnet50_moco_v2	.624
290	deit_large_imagenet_full_seed-0	.624
291	omnivore_swinT	.623
292	mobilenet_v2_1_0_224	.622
293	efficientnet_b2_imagenet_full	.619
294	yudixie_resnet18_random_0_240719	.613
295	artResNet18_1	.612
296	custom_model_cv_18_dagger_408	.612
297	inception_v1-tf	.610
298	resnet50_simclr	.602
299	mobilenet_v2_0.35_224	.602
300	mobilenet_v2_1.0_128-tf	.596
301	mobilenet_v2_1_0_160	.594
302	effnetb1_272x240	.594
303	vit_large_patch14_clip_224:openai_ft_in1k	.589
304	r3m_resnet34	.585
305	mobilenet_v2_0.5_224-tf	.582
306	resnet-18-LC_conv_init_m	.583
307	vit_huge_patch14_clip_224:laion2b_ft_in12k_in1k	.578
308	convnext_small_imagenet_1_seed-0	.577
309	vit_large_patch14_clip_224:openai_ft_in12k_in1k	.561
310	mobilenet_v2_0_5_192	.559
311	hmax	.553
312	deit_base_imagenet_full_seed-0	.551
313	vit_relpos_base_patch16_clsgap_224:sw_in1k	.550
314	r3m_resnet50_nocrop	.528
315	AT_efficientnet-b2	.526
316	AdvProp_efficientnet-b2	.525
317	inception_resnet_v2	.521
318	r3m_resnet18	.514
319	r3m_resnet50	.506
320	resnet50_pretrained_with_retinal_waves	.489
321	my-model	.484
322	mobilenet_v2_0.75_96	.481
323	alexnet_training_seed_08	.479
324	CORnetZ_CIFAR10	.422
325	mobilenet_v2_0.5_160	.398
326	mobilenet_v2_0.75_224-tf	.397
327	deit_small_imagenet_10_seed-0	.388
328	ViT-B/32	.387
329	resnet50_imagenet_1_seed-0	.325
330	alexnet_training_seed_05	.313
331	alexnet_training_seed_06	.233
332	resnet-18-LC_untrained	.233
333	resnet50-meshes-lt-100-original-pretrained	.177
334	resnet50-meshes-lt-100-original-scratch	.115
335	mobilenet_v2_0.5_128	.038
336	yudixie_resnet50_imagenet1kpret_0_240312	X
337	resnet_152_v1	X
338	mobilenet_v2_0.35_160	X
339	resnet50_ImageNet	X
340	pnasnet_large	X
341	alexnet_ks_torevert	X
342	vonegrcnn_47e	X
343	resnext101_32x48d_wsl	X
344	vonegrcnn_52e_full	X
345	vonegrcnn_47e	X
346	deit_small_imagenet_1_seed-0	X
347	nasnet_large	X
348	resnet18-supervised	X
349	resnext101_32x32d_wsl	X
350	fitvid_trained_on_physion	X
351	inception_v4	X
352	tv_efficientnet-b1	X
353	ViT_L_32_imagenet1k	X
354	regnet_y_400mf	X
355	resnext101_32x16d_wsl	X
356	densenet_201_pytorch	X
357	inception_v3_pytorch	X
358	pnasnet_large_pytorch	X
359	resnet_152_v2	X
360	resnet_101_v1	X
361	vonegrcnn_62e_nobn	X
362	openclip	X
363	resnet_101_v2	X

Benchmark bibtex

None

Ceiling

0.96.

Note that scores are relative to this ceiling.

Data: Marques2020_Ringach2002

Metric: orth_pref_ratio