Sample stimuli

sample 0 sample 1 sample 2 sample 3 sample 4 sample 5 sample 6 sample 7 sample 8 sample 9

How to use

from brainscore_vision import load_benchmark
benchmark = load_benchmark("Geirhos2021uniformnoise-error_consistency")
score = benchmark(my_model)

Model scores

Min Alignment Max Alignment

Rank

Model

Score

1
.781
2
.754
3
.743
4
.734
5
.731
6
.728
7
.723
8
.720
9
.719
10
.686
11
.685
12
.675
13
.674
14
.670
15
.661
16
.647
17
.647
18
.645
19
.639
20
.637
21
.635
22
.627
23
.627
24
.617
25
.614
26
.604
27
.601
28
.601
29
.599
30
.591
31
.590
32
.587
33
.583
34
.577
35
.576
36
.576
37
.572
38
.555
39
.537
40
.524
41
.522
42
.520
43
.520
44
.515
45
.509
46
.493
47
.493
48
.476
49
.475
50
.473
51
.462
52
.458
53
.455
54
.443
55
.441
56
.418
57
.404
58
.401
59
.386
60
.381
61
.377
62
.354
63
.348
64
.344
65
.340
66
.338
67
.330
68
.328
69
.322
70
.309
71
.309
72
.309
73
.289
74
.289
75
.285
76
.284
77
.282
78
.279
79
.272
80
.266
81
.265
82
.260
83
.257
84
.257
85
.254
86
.253
87
.250
88
.250
89
.249
90
.245
91
.244
92
.243
93
.231
94
.231
95
.224
96
.209
97
.209
98
.202
99
.196
100
.194
101
.194
102
.194
103
.194
104
.194
105
.188
106
.188
107
.183
108
.182
109
.180
110
.179
111
.174
112
.173
113
.172
114
.172
115
.166
116
.162
117
.162
118
.157
119
.157
120
.156
121
.153
122
.149
123
.144
124
.143
125
.137
126
.136
127
.134
128
.133
129
.132
130
.131
131
.130
132
.130
133
.130
134
.130
135
.130
136
.129
137
.127
138
.127
139
.125
140
.122
141
.121
142
.121
143
.121
144
.119
145
.119
146
.116
147
.112
148
.105
149
.105
150
.103
151
.102
152
.101
153
.099
154
.098
155
.097
156
.091
157
.090
158
.089
159
.089
160
.089
161
.088
162
.088
163
.088
164
.088
165
.088
166
.088
167
.088
168
.080
169
.079
170
.079
171
.078
172
.078
173
.078
174
.077
175
.076
176
.075
177
.075
178
.075
179
.073
180
.073
181
.071
182
.069
183
.068
184
.068
185
.067
186
.067
187
.067
188
.064
189
.062
190
.061
191
.061
192
.061
193
.059
194
.059
195
.058
196
.058
197
.058
198
.057
199
.057
200
.057
201
.057
202
.057
203
.056
204
.056
205
.054
206
.054
207
.050
208
.049
209
.049
210
.049
211
.049
212
.048
213
.048
214
.048
215
.048
216
.047
217
.047
218
.047
219
.046
220
.046
221
.046
222
.046
223
.046
224
.045
225
.045
226
.045
227
.044
228
.044
229
.043
230
.043
231
.041
232
.041
233
.040
234
.039
235
.039
236
.038
237
.038
238
.038
239
.038
240
.038
241
.038
242
.038
243
.038
244
.038
245
.038
246
.038
247
.038
248
.038
249
.038
250
.038
251
.038
252
.038
253
.036
254
.036
255
.036
256
.036
257
.036
258
.035
259
.034
260
.034
261
.034
262
.033
263
.033
264
.031
265
.031
266
.030
267
.028
268
.028
269
.027
270
.027
271
.026
272
.024
273
.024
274
.023
275
.022
276
.022
277
.022
278
.022
279
.021
280
.021
281
.020
282
.017
283
.016
284
.016
285
.014
286
.013
287
.012
288
.011
289
.011
290
.011
291
.010
292
.010
293
.009
294
.008
295
.006
296
.006
297
.006
298
.005
299
.005
300
.005
301
.005
302
.004
303
.004
304
.003
305
.003
306
.002
307
.001
308
.000
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455

Benchmark bibtex

@article{geirhos2021partial,
              title={Partial success in closing the gap between human and machine vision},
              author={Geirhos, Robert and Narayanappa, Kantharaju and Mitzkus, Benjamin and Thieringer, Tizian and Bethge, Matthias and Wichmann, Felix A and Brendel, Wieland},
              journal={Advances in Neural Information Processing Systems},
              volume={34},
              year={2021},
              url={https://openreview.net/forum?id=QkljT4mrfs}
        }

Ceiling

0.43.

Note that scores are relative to this ceiling.

Data: Geirhos2021uniformnoise

Metric: error_consistency