latent-research-lab commited on
Commit
6927d2c
·
verified ·
1 Parent(s): e0139c5

metadata: added Croissant metadata

Browse files
Files changed (1) hide show
  1. croissant.json +691 -0
croissant.json ADDED
@@ -0,0 +1,691 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "@context": {
3
+ "@language": "en",
4
+ "@vocab": "https://schema.org/",
5
+ "arrayShape": "cr:arrayShape",
6
+ "citeAs": "cr:citeAs",
7
+ "column": "cr:column",
8
+ "conformsTo": "dct:conformsTo",
9
+ "containedIn": "cr:containedIn",
10
+ "cr": "http://mlcommons.org/croissant/",
11
+ "data": { "@id": "cr:data", "@type": "@json" },
12
+ "dataCollection": "cr:dataCollection",
13
+ "dataType": { "@id": "cr:dataType", "@type": "@vocab" },
14
+ "dct": "http://purl.org/dc/terms/",
15
+ "extract": "cr:extract",
16
+ "field": "cr:field",
17
+ "fileProperty": "cr:fileProperty",
18
+ "fileObject": "cr:fileObject",
19
+ "fileSet": "cr:fileSet",
20
+ "format": "cr:format",
21
+ "includes": "cr:includes",
22
+ "isArray": "cr:isArray",
23
+ "isLiveDataset": "cr:isLiveDataset",
24
+ "jsonPath": "cr:jsonPath",
25
+ "key": "cr:key",
26
+ "md5": "cr:md5",
27
+ "parentField": "cr:parentField",
28
+ "path": "cr:path",
29
+ "recordSet": "cr:recordSet",
30
+ "references": "cr:references",
31
+ "regex": "cr:regex",
32
+ "repeated": "cr:repeated",
33
+ "replace": "cr:replace",
34
+ "sc": "https://schema.org/",
35
+ "separator": "cr:separator",
36
+ "source": "cr:source",
37
+ "subField": "cr:subField",
38
+ "transform": "cr:transform",
39
+ "rai": "http://mlcommons.org/croissant/rai/",
40
+ "prov": "http://www.w3.org/ns/prov#",
41
+ "dataLimitations": "rai:dataLimitations",
42
+ "dataBiases": "rai:dataBiases",
43
+ "personalSensitiveInformation": "rai:personalSensitiveInformation",
44
+ "dataUseCases": "rai:dataUseCases",
45
+ "dataSocialImpact": "rai:dataSocialImpact",
46
+ "hasSyntheticData": "rai:hasSyntheticData",
47
+ "wasDerivedFrom": "prov:wasDerivedFrom",
48
+ "wasGeneratedBy": "prov:wasGeneratedBy"
49
+ },
50
+ "@type": "sc:Dataset",
51
+ "distribution": [
52
+ {
53
+ "@type": "cr:FileObject",
54
+ "@id": "repo",
55
+ "name": "repo",
56
+ "description": "The Hugging Face git repository.",
57
+ "contentUrl": "https://huggingface.co/datasets/shadow-bench/ShadowBench/tree/refs%2Fconvert%2Fparquet",
58
+ "encodingFormat": "git+https",
59
+ "sha256": "https://github.com/mlcommons/croissant/issues/80"
60
+ },
61
+ {
62
+ "@type": "cr:FileSet",
63
+ "@id": "parquet-files-for-config-entertainment",
64
+ "containedIn": { "@id": "repo" },
65
+ "encodingFormat": "application/x-parquet",
66
+ "includes": "entertainment/*/*.parquet"
67
+ },
68
+ {
69
+ "@type": "cr:FileSet",
70
+ "@id": "parquet-files-for-config-sports",
71
+ "containedIn": { "@id": "repo" },
72
+ "encodingFormat": "application/x-parquet",
73
+ "includes": "sports/*/*.parquet"
74
+ },
75
+ {
76
+ "@type": "cr:FileSet",
77
+ "@id": "parquet-files-for-config-technology",
78
+ "containedIn": { "@id": "repo" },
79
+ "encodingFormat": "application/x-parquet",
80
+ "includes": "technology/*/*.parquet"
81
+ }
82
+ ],
83
+ "recordSet": [
84
+ {
85
+ "@type": "cr:RecordSet",
86
+ "dataType": "cr:Split",
87
+ "key": { "@id": "entertainment_splits/split_name" },
88
+ "@id": "entertainment_splits",
89
+ "name": "entertainment_splits",
90
+ "description": "Splits for the entertainment config.",
91
+ "field": [
92
+ {
93
+ "@type": "cr:Field",
94
+ "@id": "entertainment_splits/split_name",
95
+ "dataType": "sc:Text"
96
+ }
97
+ ],
98
+ "data": [
99
+ { "entertainment_splits/split_name": "upper_shadow" },
100
+ { "entertainment_splits/split_name": "lower_shadow" },
101
+ {
102
+ "entertainment_splits/split_name": "upper_shadow_controlled"
103
+ },
104
+ {
105
+ "entertainment_splits/split_name": "lower_shadow_controlled"
106
+ },
107
+ { "entertainment_splits/split_name": "upper_direct" },
108
+ { "entertainment_splits/split_name": "lower_direct" }
109
+ ]
110
+ },
111
+ {
112
+ "@type": "cr:RecordSet",
113
+ "@id": "entertainment",
114
+ "description": "shadow-bench/ShadowBench - 'entertainment' subset\n\nAdditional information:\n- 6 splits: upper_shadow, lower_shadow, upper_shadow_controlled, lower_shadow_controlled, upper_direct, lower_direct",
115
+ "field": [
116
+ {
117
+ "@type": "cr:Field",
118
+ "@id": "entertainment/split",
119
+ "dataType": "sc:Text",
120
+ "source": {
121
+ "fileSet": {
122
+ "@id": "parquet-files-for-config-entertainment"
123
+ },
124
+ "extract": { "fileProperty": "fullpath" },
125
+ "transform": {
126
+ "regex": "entertainment/(?:partial-)?(upper_shadow|lower_shadow|upper_shadow_controlled|lower_shadow_controlled|upper_direct|lower_direct)/.+parquet$"
127
+ }
128
+ },
129
+ "references": {
130
+ "field": { "@id": "entertainment_splits/split_name" }
131
+ }
132
+ },
133
+ {
134
+ "@type": "cr:Field",
135
+ "@id": "entertainment/entity",
136
+ "dataType": "sc:Text",
137
+ "source": {
138
+ "fileSet": {
139
+ "@id": "parquet-files-for-config-entertainment"
140
+ },
141
+ "extract": { "column": "entity" }
142
+ }
143
+ },
144
+ {
145
+ "@type": "cr:Field",
146
+ "@id": "entertainment/question",
147
+ "dataType": "sc:Text",
148
+ "source": {
149
+ "fileSet": {
150
+ "@id": "parquet-files-for-config-entertainment"
151
+ },
152
+ "extract": { "column": "question" }
153
+ }
154
+ },
155
+ {
156
+ "@type": "cr:Field",
157
+ "@id": "entertainment/choices",
158
+ "subField": [
159
+ {
160
+ "@type": "cr:Field",
161
+ "@id": "entertainment/choices/A",
162
+ "dataType": "sc:Text",
163
+ "source": {
164
+ "fileSet": {
165
+ "@id": "parquet-files-for-config-entertainment"
166
+ },
167
+ "extract": { "column": "choices" },
168
+ "transform": { "jsonPath": "A" }
169
+ }
170
+ },
171
+ {
172
+ "@type": "cr:Field",
173
+ "@id": "entertainment/choices/B",
174
+ "dataType": "sc:Text",
175
+ "source": {
176
+ "fileSet": {
177
+ "@id": "parquet-files-for-config-entertainment"
178
+ },
179
+ "extract": { "column": "choices" },
180
+ "transform": { "jsonPath": "B" }
181
+ }
182
+ },
183
+ {
184
+ "@type": "cr:Field",
185
+ "@id": "entertainment/choices/C",
186
+ "dataType": "sc:Text",
187
+ "source": {
188
+ "fileSet": {
189
+ "@id": "parquet-files-for-config-entertainment"
190
+ },
191
+ "extract": { "column": "choices" },
192
+ "transform": { "jsonPath": "C" }
193
+ }
194
+ },
195
+ {
196
+ "@type": "cr:Field",
197
+ "@id": "entertainment/choices/D",
198
+ "dataType": "sc:Text",
199
+ "source": {
200
+ "fileSet": {
201
+ "@id": "parquet-files-for-config-entertainment"
202
+ },
203
+ "extract": { "column": "choices" },
204
+ "transform": { "jsonPath": "D" }
205
+ }
206
+ }
207
+ ]
208
+ },
209
+ {
210
+ "@type": "cr:Field",
211
+ "@id": "entertainment/answer",
212
+ "dataType": "sc:Text",
213
+ "source": {
214
+ "fileSet": {
215
+ "@id": "parquet-files-for-config-entertainment"
216
+ },
217
+ "extract": { "column": "answer" }
218
+ }
219
+ },
220
+ {
221
+ "@type": "cr:Field",
222
+ "@id": "entertainment/metadata",
223
+ "subField": [
224
+ {
225
+ "@type": "cr:Field",
226
+ "@id": "entertainment/metadata/A_0",
227
+ "dataType": "sc:Text",
228
+ "source": {
229
+ "fileSet": {
230
+ "@id": "parquet-files-for-config-entertainment"
231
+ },
232
+ "extract": { "column": "metadata" },
233
+ "transform": { "jsonPath": "A" }
234
+ }
235
+ },
236
+ {
237
+ "@type": "cr:Field",
238
+ "@id": "entertainment/metadata/B_0",
239
+ "dataType": "sc:Text",
240
+ "source": {
241
+ "fileSet": {
242
+ "@id": "parquet-files-for-config-entertainment"
243
+ },
244
+ "extract": { "column": "metadata" },
245
+ "transform": { "jsonPath": "B" }
246
+ }
247
+ },
248
+ {
249
+ "@type": "cr:Field",
250
+ "@id": "entertainment/metadata/C_0",
251
+ "dataType": "sc:Text",
252
+ "source": {
253
+ "fileSet": {
254
+ "@id": "parquet-files-for-config-entertainment"
255
+ },
256
+ "extract": { "column": "metadata" },
257
+ "transform": { "jsonPath": "C" }
258
+ }
259
+ },
260
+ {
261
+ "@type": "cr:Field",
262
+ "@id": "entertainment/metadata/D_0",
263
+ "dataType": "sc:Text",
264
+ "source": {
265
+ "fileSet": {
266
+ "@id": "parquet-files-for-config-entertainment"
267
+ },
268
+ "extract": { "column": "metadata" },
269
+ "transform": { "jsonPath": "D" }
270
+ }
271
+ }
272
+ ]
273
+ }
274
+ ]
275
+ },
276
+ {
277
+ "@type": "cr:RecordSet",
278
+ "dataType": "cr:Split",
279
+ "key": { "@id": "sports_splits/split_name" },
280
+ "@id": "sports_splits",
281
+ "name": "sports_splits",
282
+ "description": "Splits for the sports config.",
283
+ "field": [
284
+ {
285
+ "@type": "cr:Field",
286
+ "@id": "sports_splits/split_name",
287
+ "dataType": "sc:Text"
288
+ }
289
+ ],
290
+ "data": [
291
+ { "sports_splits/split_name": "upper_shadow" },
292
+ { "sports_splits/split_name": "lower_shadow" },
293
+ { "sports_splits/split_name": "upper_shadow_controlled" },
294
+ { "sports_splits/split_name": "lower_shadow_controlled" },
295
+ { "sports_splits/split_name": "upper_direct" },
296
+ { "sports_splits/split_name": "lower_direct" }
297
+ ]
298
+ },
299
+ {
300
+ "@type": "cr:RecordSet",
301
+ "@id": "sports",
302
+ "description": "shadow-bench/ShadowBench - 'sports' subset\n\nAdditional information:\n- 6 splits: upper_shadow, lower_shadow, upper_shadow_controlled, lower_shadow_controlled, upper_direct, lower_direct",
303
+ "field": [
304
+ {
305
+ "@type": "cr:Field",
306
+ "@id": "sports/split",
307
+ "dataType": "sc:Text",
308
+ "source": {
309
+ "fileSet": { "@id": "parquet-files-for-config-sports" },
310
+ "extract": { "fileProperty": "fullpath" },
311
+ "transform": {
312
+ "regex": "sports/(?:partial-)?(upper_shadow|lower_shadow|upper_shadow_controlled|lower_shadow_controlled|upper_direct|lower_direct)/.+parquet$"
313
+ }
314
+ },
315
+ "references": {
316
+ "field": { "@id": "sports_splits/split_name" }
317
+ }
318
+ },
319
+ {
320
+ "@type": "cr:Field",
321
+ "@id": "sports/entity",
322
+ "dataType": "sc:Text",
323
+ "source": {
324
+ "fileSet": { "@id": "parquet-files-for-config-sports" },
325
+ "extract": { "column": "entity" }
326
+ }
327
+ },
328
+ {
329
+ "@type": "cr:Field",
330
+ "@id": "sports/question",
331
+ "dataType": "sc:Text",
332
+ "source": {
333
+ "fileSet": { "@id": "parquet-files-for-config-sports" },
334
+ "extract": { "column": "question" }
335
+ }
336
+ },
337
+ {
338
+ "@type": "cr:Field",
339
+ "@id": "sports/choices",
340
+ "subField": [
341
+ {
342
+ "@type": "cr:Field",
343
+ "@id": "sports/choices/A_0_0",
344
+ "dataType": "sc:Text",
345
+ "source": {
346
+ "fileSet": {
347
+ "@id": "parquet-files-for-config-sports"
348
+ },
349
+ "extract": { "column": "choices" },
350
+ "transform": { "jsonPath": "A" }
351
+ }
352
+ },
353
+ {
354
+ "@type": "cr:Field",
355
+ "@id": "sports/choices/B_0_0",
356
+ "dataType": "sc:Text",
357
+ "source": {
358
+ "fileSet": {
359
+ "@id": "parquet-files-for-config-sports"
360
+ },
361
+ "extract": { "column": "choices" },
362
+ "transform": { "jsonPath": "B" }
363
+ }
364
+ },
365
+ {
366
+ "@type": "cr:Field",
367
+ "@id": "sports/choices/C_0_0",
368
+ "dataType": "sc:Text",
369
+ "source": {
370
+ "fileSet": {
371
+ "@id": "parquet-files-for-config-sports"
372
+ },
373
+ "extract": { "column": "choices" },
374
+ "transform": { "jsonPath": "C" }
375
+ }
376
+ },
377
+ {
378
+ "@type": "cr:Field",
379
+ "@id": "sports/choices/D_0_0",
380
+ "dataType": "sc:Text",
381
+ "source": {
382
+ "fileSet": {
383
+ "@id": "parquet-files-for-config-sports"
384
+ },
385
+ "extract": { "column": "choices" },
386
+ "transform": { "jsonPath": "D" }
387
+ }
388
+ }
389
+ ]
390
+ },
391
+ {
392
+ "@type": "cr:Field",
393
+ "@id": "sports/answer",
394
+ "dataType": "sc:Text",
395
+ "source": {
396
+ "fileSet": { "@id": "parquet-files-for-config-sports" },
397
+ "extract": { "column": "answer" }
398
+ }
399
+ },
400
+ {
401
+ "@type": "cr:Field",
402
+ "@id": "sports/metadata",
403
+ "subField": [
404
+ {
405
+ "@type": "cr:Field",
406
+ "@id": "sports/metadata/A_0_0_0",
407
+ "dataType": "sc:Text",
408
+ "source": {
409
+ "fileSet": {
410
+ "@id": "parquet-files-for-config-sports"
411
+ },
412
+ "extract": { "column": "metadata" },
413
+ "transform": { "jsonPath": "A" }
414
+ }
415
+ },
416
+ {
417
+ "@type": "cr:Field",
418
+ "@id": "sports/metadata/B_0_0_0",
419
+ "dataType": "sc:Text",
420
+ "source": {
421
+ "fileSet": {
422
+ "@id": "parquet-files-for-config-sports"
423
+ },
424
+ "extract": { "column": "metadata" },
425
+ "transform": { "jsonPath": "B" }
426
+ }
427
+ },
428
+ {
429
+ "@type": "cr:Field",
430
+ "@id": "sports/metadata/C_0_0_0",
431
+ "dataType": "sc:Text",
432
+ "source": {
433
+ "fileSet": {
434
+ "@id": "parquet-files-for-config-sports"
435
+ },
436
+ "extract": { "column": "metadata" },
437
+ "transform": { "jsonPath": "C" }
438
+ }
439
+ },
440
+ {
441
+ "@type": "cr:Field",
442
+ "@id": "sports/metadata/D_0_0_0",
443
+ "dataType": "sc:Text",
444
+ "source": {
445
+ "fileSet": {
446
+ "@id": "parquet-files-for-config-sports"
447
+ },
448
+ "extract": { "column": "metadata" },
449
+ "transform": { "jsonPath": "D" }
450
+ }
451
+ }
452
+ ]
453
+ }
454
+ ]
455
+ },
456
+ {
457
+ "@type": "cr:RecordSet",
458
+ "dataType": "cr:Split",
459
+ "key": { "@id": "technology_splits/split_name" },
460
+ "@id": "technology_splits",
461
+ "name": "technology_splits",
462
+ "description": "Splits for the technology config.",
463
+ "field": [
464
+ {
465
+ "@type": "cr:Field",
466
+ "@id": "technology_splits/split_name",
467
+ "dataType": "sc:Text"
468
+ }
469
+ ],
470
+ "data": [
471
+ { "technology_splits/split_name": "upper_shadow" },
472
+ { "technology_splits/split_name": "lower_shadow" },
473
+ { "technology_splits/split_name": "upper_shadow_controlled" },
474
+ { "technology_splits/split_name": "lower_shadow_controlled" },
475
+ { "technology_splits/split_name": "upper_direct" },
476
+ { "technology_splits/split_name": "lower_direct" }
477
+ ]
478
+ },
479
+ {
480
+ "@type": "cr:RecordSet",
481
+ "@id": "technology",
482
+ "description": "shadow-bench/ShadowBench - 'technology' subset\n\nAdditional information:\n- 6 splits: upper_shadow, lower_shadow, upper_shadow_controlled, lower_shadow_controlled, upper_direct, lower_direct",
483
+ "field": [
484
+ {
485
+ "@type": "cr:Field",
486
+ "@id": "technology/split",
487
+ "dataType": "sc:Text",
488
+ "source": {
489
+ "fileSet": {
490
+ "@id": "parquet-files-for-config-technology"
491
+ },
492
+ "extract": { "fileProperty": "fullpath" },
493
+ "transform": {
494
+ "regex": "technology/(?:partial-)?(upper_shadow|lower_shadow|upper_shadow_controlled|lower_shadow_controlled|upper_direct|lower_direct)/.+parquet$"
495
+ }
496
+ },
497
+ "references": {
498
+ "field": { "@id": "technology_splits/split_name" }
499
+ }
500
+ },
501
+ {
502
+ "@type": "cr:Field",
503
+ "@id": "technology/entity",
504
+ "dataType": "sc:Text",
505
+ "source": {
506
+ "fileSet": {
507
+ "@id": "parquet-files-for-config-technology"
508
+ },
509
+ "extract": { "column": "entity" }
510
+ }
511
+ },
512
+ {
513
+ "@type": "cr:Field",
514
+ "@id": "technology/question",
515
+ "dataType": "sc:Text",
516
+ "source": {
517
+ "fileSet": {
518
+ "@id": "parquet-files-for-config-technology"
519
+ },
520
+ "extract": { "column": "question" }
521
+ }
522
+ },
523
+ {
524
+ "@type": "cr:Field",
525
+ "@id": "technology/choices",
526
+ "subField": [
527
+ {
528
+ "@type": "cr:Field",
529
+ "@id": "technology/choices/A_0_0_0_0",
530
+ "dataType": "sc:Text",
531
+ "source": {
532
+ "fileSet": {
533
+ "@id": "parquet-files-for-config-technology"
534
+ },
535
+ "extract": { "column": "choices" },
536
+ "transform": { "jsonPath": "A" }
537
+ }
538
+ },
539
+ {
540
+ "@type": "cr:Field",
541
+ "@id": "technology/choices/B_0_0_0_0",
542
+ "dataType": "sc:Text",
543
+ "source": {
544
+ "fileSet": {
545
+ "@id": "parquet-files-for-config-technology"
546
+ },
547
+ "extract": { "column": "choices" },
548
+ "transform": { "jsonPath": "B" }
549
+ }
550
+ },
551
+ {
552
+ "@type": "cr:Field",
553
+ "@id": "technology/choices/C_0_0_0_0",
554
+ "dataType": "sc:Text",
555
+ "source": {
556
+ "fileSet": {
557
+ "@id": "parquet-files-for-config-technology"
558
+ },
559
+ "extract": { "column": "choices" },
560
+ "transform": { "jsonPath": "C" }
561
+ }
562
+ },
563
+ {
564
+ "@type": "cr:Field",
565
+ "@id": "technology/choices/D_0_0_0_0",
566
+ "dataType": "sc:Text",
567
+ "source": {
568
+ "fileSet": {
569
+ "@id": "parquet-files-for-config-technology"
570
+ },
571
+ "extract": { "column": "choices" },
572
+ "transform": { "jsonPath": "D" }
573
+ }
574
+ }
575
+ ]
576
+ },
577
+ {
578
+ "@type": "cr:Field",
579
+ "@id": "technology/answer",
580
+ "dataType": "sc:Text",
581
+ "source": {
582
+ "fileSet": {
583
+ "@id": "parquet-files-for-config-technology"
584
+ },
585
+ "extract": { "column": "answer" }
586
+ }
587
+ },
588
+ {
589
+ "@type": "cr:Field",
590
+ "@id": "technology/metadata",
591
+ "subField": [
592
+ {
593
+ "@type": "cr:Field",
594
+ "@id": "technology/metadata/A_0_0_0_0_0",
595
+ "dataType": "sc:Text",
596
+ "source": {
597
+ "fileSet": {
598
+ "@id": "parquet-files-for-config-technology"
599
+ },
600
+ "extract": { "column": "metadata" },
601
+ "transform": { "jsonPath": "A" }
602
+ }
603
+ },
604
+ {
605
+ "@type": "cr:Field",
606
+ "@id": "technology/metadata/B_0_0_0_0_0",
607
+ "dataType": "sc:Text",
608
+ "source": {
609
+ "fileSet": {
610
+ "@id": "parquet-files-for-config-technology"
611
+ },
612
+ "extract": { "column": "metadata" },
613
+ "transform": { "jsonPath": "B" }
614
+ }
615
+ },
616
+ {
617
+ "@type": "cr:Field",
618
+ "@id": "technology/metadata/C_0_0_0_0_0",
619
+ "dataType": "sc:Text",
620
+ "source": {
621
+ "fileSet": {
622
+ "@id": "parquet-files-for-config-technology"
623
+ },
624
+ "extract": { "column": "metadata" },
625
+ "transform": { "jsonPath": "C" }
626
+ }
627
+ },
628
+ {
629
+ "@type": "cr:Field",
630
+ "@id": "technology/metadata/D_0_0_0_0_0",
631
+ "dataType": "sc:Text",
632
+ "source": {
633
+ "fileSet": {
634
+ "@id": "parquet-files-for-config-technology"
635
+ },
636
+ "extract": { "column": "metadata" },
637
+ "transform": { "jsonPath": "D" }
638
+ }
639
+ }
640
+ ]
641
+ }
642
+ ]
643
+ }
644
+ ],
645
+ "conformsTo": "http://mlcommons.org/croissant/1.1",
646
+ "name": "ShadowBench",
647
+ "description": "\n\t\n\t\t\n\t\tShadowBench: A Hardened Benchmark for Latent Entity Association\n\t\n\nShadowBench is a diagnostic framework designed to evaluate the \"Shadow Knowledge\" of Large Language Models (LLMs). While traditional benchmarks measure factual recall using explicit entity names (e.g., \"Elon Musk\"), ShadowBench evaluates whether a model can navigate its internal knowledge graph when these lexical anchors are removed.\n\n\t\n\t\t\n\t\n\t\n\t\tDataset Summary\n\t\n\nThe core task in ShadowBench is Dual-Trait Association… See the full description on the dataset page: https://huggingface.co/datasets/shadow-bench/ShadowBench.",
648
+ "alternateName": ["shadow-bench/ShadowBench"],
649
+ "creator": {
650
+ "@type": "Organization",
651
+ "name": "ShadowBench",
652
+ "url": "https://huggingface.co/shadow-bench"
653
+ },
654
+ "keywords": [
655
+ "question-answering",
656
+ "multiple-choice",
657
+ "English",
658
+ "cc-by-sa-4.0",
659
+ "10K - 100K",
660
+ "parquet",
661
+ "Text",
662
+ "Datasets",
663
+ "pandas",
664
+ "Polars",
665
+ "Croissant",
666
+ "🇺🇸 Region: US",
667
+ "knowledge-probing",
668
+ "llm-evaluation",
669
+ "entity-resolution",
670
+ "machine-unlearning"
671
+ ],
672
+ "license": "https://choosealicense.com/licenses/cc-by-sa-4.0/",
673
+ "url": "https://huggingface.co/datasets/shadow-bench/ShadowBench",
674
+ "rai:dataLimitations": "The dataset is limited to English-language Wikipedia records and three specific domains: Technology, Sports (Tennis), and Entertainment (Actors). It is restricted to the January 1, 2023 Wikipedia snapshot and may not reflect real-world changes or entities emerging after this date.",
675
+ "rai:dataBiases": "The dataset contains an intentional popularity bias used to evaluate the 'Knowledge Cliff' phenomenon. There is a potential demographic skew reflecting Wikipedia's historical representation patterns, primarily favoring Western-centric public figures.",
676
+ "rai:personalSensitiveInformation": "The dataset contains names and biographical facts of public figures (celebrities, tech executives, athletes). It does not contain private PII, non-public contact information, medical data, or financial records.",
677
+ "rai:dataUseCases": "Intended for the evaluation of latent knowledge representation in LLMs and the diagnostic testing of Machine Unlearning robustness. It is not intended for use as a training or fine-tuning corpus.",
678
+ "rai:dataSocialImpact": "ShadowBench aims to improve AI safety by exposing vulnerabilities in privacy-preserving unlearning methods. It highlights the risk of 'Superficial Forgetting' where sensitive information remains retrievable via latent associations.",
679
+ "rai:hasSyntheticData": false,
680
+ "prov:wasDerivedFrom": [
681
+ {
682
+ "@type": "sc:Dataset",
683
+ "url": "https://en.wikipedia.org/",
684
+ "description": "English Wikipedia Snapshot (January 1, 2023)"
685
+ }
686
+ ],
687
+ "prov:wasGeneratedBy": {
688
+ "@type": "prov:Activity",
689
+ "description": "The dataset was generated through a multi-stage pipeline: (1) Breadth-First Search (BFS) Wikipedia category traversal, (2) Named Entity Recognition (NER) for factual density filtering, (3) Lexical and syntactic anonymization of pronouns and entity names, (4) Generational Proximity Filtering for distractor matching, and (5) a human-in-the-loop audit for bijective mapping and factual accuracy."
690
+ }
691
+ }