amirali1985 commited on
Commit
a13566b
·
verified ·
1 Parent(s): 04b241e

Upload folder using huggingface_hub

Browse files
interp_results/add_sub_sorl_v1_abs30_100K/causal_verification.json ADDED
@@ -0,0 +1,335 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model": "add_sub_sorl_v1_abs30_100K",
3
+ "K": 4,
4
+ "n_examples": 1400,
5
+ "baseline": {
6
+ "correct": 1352,
7
+ "total": 1400,
8
+ "accuracy": 0.9657142857142857
9
+ },
10
+ "knockout": {
11
+ "correct": 0,
12
+ "total": 1400,
13
+ "accuracy": 0.0
14
+ },
15
+ "shuffle": {
16
+ "correct": 1297,
17
+ "total": 1400,
18
+ "accuracy": 0.9264285714285714
19
+ },
20
+ "random": {
21
+ "correct": 1253,
22
+ "total": 1400,
23
+ "accuracy": 0.895
24
+ },
25
+ "per_split": {
26
+ "add_S0": {
27
+ "baseline": {
28
+ "accuracy": 1.0
29
+ },
30
+ "knockout": {
31
+ "accuracy": 0.0
32
+ },
33
+ "shuffle": {
34
+ "accuracy": 1.0
35
+ },
36
+ "random": {
37
+ "accuracy": 1.0
38
+ }
39
+ },
40
+ "add_S1": {
41
+ "baseline": {
42
+ "accuracy": 1.0
43
+ },
44
+ "knockout": {
45
+ "accuracy": 0.0
46
+ },
47
+ "shuffle": {
48
+ "accuracy": 0.96
49
+ },
50
+ "random": {
51
+ "accuracy": 1.0
52
+ }
53
+ },
54
+ "add_S2": {
55
+ "baseline": {
56
+ "accuracy": 1.0
57
+ },
58
+ "knockout": {
59
+ "accuracy": 0.0
60
+ },
61
+ "shuffle": {
62
+ "accuracy": 1.0
63
+ },
64
+ "random": {
65
+ "accuracy": 1.0
66
+ }
67
+ },
68
+ "add_S3": {
69
+ "baseline": {
70
+ "accuracy": 1.0
71
+ },
72
+ "knockout": {
73
+ "accuracy": 0.0
74
+ },
75
+ "shuffle": {
76
+ "accuracy": 0.94
77
+ },
78
+ "random": {
79
+ "accuracy": 0.94
80
+ }
81
+ },
82
+ "add_S4": {
83
+ "baseline": {
84
+ "accuracy": 0.96
85
+ },
86
+ "knockout": {
87
+ "accuracy": 0.0
88
+ },
89
+ "shuffle": {
90
+ "accuracy": 0.9
91
+ },
92
+ "random": {
93
+ "accuracy": 0.72
94
+ }
95
+ },
96
+ "add_S5": {
97
+ "baseline": {
98
+ "accuracy": 0.72
99
+ },
100
+ "knockout": {
101
+ "accuracy": 0.0
102
+ },
103
+ "shuffle": {
104
+ "accuracy": 0.4
105
+ },
106
+ "random": {
107
+ "accuracy": 0.32
108
+ }
109
+ },
110
+ "add_S6": {
111
+ "baseline": {
112
+ "accuracy": 0.98
113
+ },
114
+ "knockout": {
115
+ "accuracy": 0.0
116
+ },
117
+ "shuffle": {
118
+ "accuracy": 0.68
119
+ },
120
+ "random": {
121
+ "accuracy": 0.5
122
+ }
123
+ },
124
+ "add_random": {
125
+ "baseline": {
126
+ "accuracy": 1.0
127
+ },
128
+ "knockout": {
129
+ "accuracy": 0.0
130
+ },
131
+ "shuffle": {
132
+ "accuracy": 0.99
133
+ },
134
+ "random": {
135
+ "accuracy": 0.985
136
+ }
137
+ },
138
+ "add_C3": {
139
+ "baseline": {
140
+ "accuracy": 0.98
141
+ },
142
+ "knockout": {
143
+ "accuracy": 0.0
144
+ },
145
+ "shuffle": {
146
+ "accuracy": 0.9
147
+ },
148
+ "random": {
149
+ "accuracy": 0.92
150
+ }
151
+ },
152
+ "add_C4": {
153
+ "baseline": {
154
+ "accuracy": 0.98
155
+ },
156
+ "knockout": {
157
+ "accuracy": 0.0
158
+ },
159
+ "shuffle": {
160
+ "accuracy": 0.92
161
+ },
162
+ "random": {
163
+ "accuracy": 0.92
164
+ }
165
+ },
166
+ "add_C5": {
167
+ "baseline": {
168
+ "accuracy": 1.0
169
+ },
170
+ "knockout": {
171
+ "accuracy": 0.0
172
+ },
173
+ "shuffle": {
174
+ "accuracy": 0.96
175
+ },
176
+ "random": {
177
+ "accuracy": 1.0
178
+ }
179
+ },
180
+ "add_C6": {
181
+ "baseline": {
182
+ "accuracy": 0.96
183
+ },
184
+ "knockout": {
185
+ "accuracy": 0.0
186
+ },
187
+ "shuffle": {
188
+ "accuracy": 0.92
189
+ },
190
+ "random": {
191
+ "accuracy": 0.9
192
+ }
193
+ },
194
+ "sub_M0": {
195
+ "baseline": {
196
+ "accuracy": 1.0
197
+ },
198
+ "knockout": {
199
+ "accuracy": 0.0
200
+ },
201
+ "shuffle": {
202
+ "accuracy": 1.0
203
+ },
204
+ "random": {
205
+ "accuracy": 1.0
206
+ }
207
+ },
208
+ "sub_M1": {
209
+ "baseline": {
210
+ "accuracy": 1.0
211
+ },
212
+ "knockout": {
213
+ "accuracy": 0.0
214
+ },
215
+ "shuffle": {
216
+ "accuracy": 1.0
217
+ },
218
+ "random": {
219
+ "accuracy": 1.0
220
+ }
221
+ },
222
+ "sub_M2": {
223
+ "baseline": {
224
+ "accuracy": 1.0
225
+ },
226
+ "knockout": {
227
+ "accuracy": 0.0
228
+ },
229
+ "shuffle": {
230
+ "accuracy": 1.0
231
+ },
232
+ "random": {
233
+ "accuracy": 0.98
234
+ }
235
+ },
236
+ "sub_M3": {
237
+ "baseline": {
238
+ "accuracy": 0.98
239
+ },
240
+ "knockout": {
241
+ "accuracy": 0.0
242
+ },
243
+ "shuffle": {
244
+ "accuracy": 0.98
245
+ },
246
+ "random": {
247
+ "accuracy": 0.96
248
+ }
249
+ },
250
+ "sub_M4": {
251
+ "baseline": {
252
+ "accuracy": 0.84
253
+ },
254
+ "knockout": {
255
+ "accuracy": 0.0
256
+ },
257
+ "shuffle": {
258
+ "accuracy": 0.74
259
+ },
260
+ "random": {
261
+ "accuracy": 0.64
262
+ }
263
+ },
264
+ "sub_M5": {
265
+ "baseline": {
266
+ "accuracy": 0.68
267
+ },
268
+ "knockout": {
269
+ "accuracy": 0.0
270
+ },
271
+ "shuffle": {
272
+ "accuracy": 0.74
273
+ },
274
+ "random": {
275
+ "accuracy": 0.6
276
+ }
277
+ },
278
+ "sub_random": {
279
+ "baseline": {
280
+ "accuracy": 0.995
281
+ },
282
+ "knockout": {
283
+ "accuracy": 0.0
284
+ },
285
+ "shuffle": {
286
+ "accuracy": 0.995
287
+ },
288
+ "random": {
289
+ "accuracy": 0.975
290
+ }
291
+ },
292
+ "sub_B3": {
293
+ "baseline": {
294
+ "accuracy": 0.98
295
+ },
296
+ "knockout": {
297
+ "accuracy": 0.0
298
+ },
299
+ "shuffle": {
300
+ "accuracy": 1.0
301
+ },
302
+ "random": {
303
+ "accuracy": 0.94
304
+ }
305
+ },
306
+ "sub_B4": {
307
+ "baseline": {
308
+ "accuracy": 1.0
309
+ },
310
+ "knockout": {
311
+ "accuracy": 0.0
312
+ },
313
+ "shuffle": {
314
+ "accuracy": 0.96
315
+ },
316
+ "random": {
317
+ "accuracy": 0.96
318
+ }
319
+ },
320
+ "sub_B5": {
321
+ "baseline": {
322
+ "accuracy": 1.0
323
+ },
324
+ "knockout": {
325
+ "accuracy": 0.0
326
+ },
327
+ "shuffle": {
328
+ "accuracy": 1.0
329
+ },
330
+ "random": {
331
+ "accuracy": 0.92
332
+ }
333
+ }
334
+ }
335
+ }