icecubesaad commited on
Commit
67e1675
·
verified ·
1 Parent(s): a18f8c5

Upload eval_results/grpo.json with huggingface_hub

Browse files
Files changed (1) hide show
  1. eval_results/grpo.json +1308 -0
eval_results/grpo.json ADDED
@@ -0,0 +1,1308 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "schema": "ladderbench.report/v1",
3
+ "generated_at": "2026-10-01T12:31:43.095097+00:00",
4
+ "ladderbench_version": "0.1.0",
5
+ "model": "grpo",
6
+ "endpoint": "vllm-offline",
7
+ "probe_set": "all@v1",
8
+ "levels_ordered": [
9
+ "xhigh",
10
+ "medium",
11
+ "low"
12
+ ],
13
+ "summary": {
14
+ "levels": [
15
+ "xhigh",
16
+ "medium",
17
+ "low"
18
+ ],
19
+ "accuracy": [
20
+ 0.8970588235294118,
21
+ 0.8823529411764706,
22
+ 0.8970588235294118
23
+ ],
24
+ "median_thinking_tokens": [
25
+ 60.0,
26
+ 47.5,
27
+ 45.0
28
+ ],
29
+ "spearman_tokens_vs_effort": 0.9999999999999998,
30
+ "spearman_accuracy_vs_effort": 0.0,
31
+ "top_level": {
32
+ "level": "xhigh",
33
+ "accuracy": 0.8970588235294118,
34
+ "median_thinking_tokens": 60.0,
35
+ "empty_trace_rate": 0.0
36
+ },
37
+ "token_spread_max_over_min": 1.3333333333333333
38
+ },
39
+ "ladder_integrity": "healthy",
40
+ "findings": [],
41
+ "calibration_delta": [
42
+ {
43
+ "level": "xhigh",
44
+ "d_accuracy": 0.0147,
45
+ "d_thinking_tokens": -5.0
46
+ },
47
+ {
48
+ "level": "medium",
49
+ "d_accuracy": 0.0,
50
+ "d_thinking_tokens": 0.0
51
+ },
52
+ {
53
+ "level": "low",
54
+ "d_accuracy": 0.0,
55
+ "d_thinking_tokens": 0.0
56
+ }
57
+ ],
58
+ "per_level": {
59
+ "xhigh": {
60
+ "n": 68,
61
+ "accuracy": 0.8971,
62
+ "median_thinking_tokens": 60.0,
63
+ "empty_trace_rate": 0.0,
64
+ "responses": [
65
+ {
66
+ "id": "e01",
67
+ "correct": true,
68
+ "predicted": "43",
69
+ "trace_chars": 121
70
+ },
71
+ {
72
+ "id": "e02",
73
+ "correct": true,
74
+ "predicted": "62",
75
+ "trace_chars": 208
76
+ },
77
+ {
78
+ "id": "e03",
79
+ "correct": true,
80
+ "predicted": "12",
81
+ "trace_chars": 205
82
+ },
83
+ {
84
+ "id": "e04",
85
+ "correct": true,
86
+ "predicted": "12",
87
+ "trace_chars": 110
88
+ },
89
+ {
90
+ "id": "e05",
91
+ "correct": true,
92
+ "predicted": "150",
93
+ "trace_chars": 96
94
+ },
95
+ {
96
+ "id": "e06",
97
+ "correct": true,
98
+ "predicted": "5.75",
99
+ "trace_chars": 269
100
+ },
101
+ {
102
+ "id": "e07",
103
+ "correct": true,
104
+ "predicted": "33",
105
+ "trace_chars": 112
106
+ },
107
+ {
108
+ "id": "e08",
109
+ "correct": true,
110
+ "predicted": "1024",
111
+ "trace_chars": 150
112
+ },
113
+ {
114
+ "id": "e09",
115
+ "correct": true,
116
+ "predicted": "210",
117
+ "trace_chars": 208
118
+ },
119
+ {
120
+ "id": "e10",
121
+ "correct": true,
122
+ "predicted": "495",
123
+ "trace_chars": 205
124
+ },
125
+ {
126
+ "id": "e11",
127
+ "correct": true,
128
+ "predicted": "28",
129
+ "trace_chars": 302
130
+ },
131
+ {
132
+ "id": "e12",
133
+ "correct": true,
134
+ "predicted": "63",
135
+ "trace_chars": 205
136
+ },
137
+ {
138
+ "id": "e13",
139
+ "correct": true,
140
+ "predicted": "64",
141
+ "trace_chars": 253
142
+ },
143
+ {
144
+ "id": "e14",
145
+ "correct": true,
146
+ "predicted": "144",
147
+ "trace_chars": 201
148
+ },
149
+ {
150
+ "id": "e15",
151
+ "correct": true,
152
+ "predicted": "7",
153
+ "trace_chars": 131
154
+ },
155
+ {
156
+ "id": "e16",
157
+ "correct": true,
158
+ "predicted": "8",
159
+ "trace_chars": 86
160
+ },
161
+ {
162
+ "id": "e17",
163
+ "correct": true,
164
+ "predicted": "12",
165
+ "trace_chars": 229
166
+ },
167
+ {
168
+ "id": "e18",
169
+ "correct": true,
170
+ "predicted": "100",
171
+ "trace_chars": 113
172
+ },
173
+ {
174
+ "id": "e19",
175
+ "correct": true,
176
+ "predicted": "90",
177
+ "trace_chars": 227
178
+ },
179
+ {
180
+ "id": "e20",
181
+ "correct": true,
182
+ "predicted": "125",
183
+ "trace_chars": 197
184
+ },
185
+ {
186
+ "id": "m01",
187
+ "correct": true,
188
+ "predicted": "1081",
189
+ "trace_chars": 105
190
+ },
191
+ {
192
+ "id": "m02",
193
+ "correct": true,
194
+ "predicted": "32768",
195
+ "trace_chars": 202
196
+ },
197
+ {
198
+ "id": "m03",
199
+ "correct": true,
200
+ "predicted": "5050",
201
+ "trace_chars": 228
202
+ },
203
+ {
204
+ "id": "m04",
205
+ "correct": true,
206
+ "predicted": "5",
207
+ "trace_chars": 317
208
+ },
209
+ {
210
+ "id": "m05",
211
+ "correct": true,
212
+ "predicted": "36",
213
+ "trace_chars": 340
214
+ },
215
+ {
216
+ "id": "m06",
217
+ "correct": true,
218
+ "predicted": "12",
219
+ "trace_chars": 185
220
+ },
221
+ {
222
+ "id": "m07",
223
+ "correct": true,
224
+ "predicted": "300",
225
+ "trace_chars": 301
226
+ },
227
+ {
228
+ "id": "m08",
229
+ "correct": true,
230
+ "predicted": "204",
231
+ "trace_chars": 432
232
+ },
233
+ {
234
+ "id": "m09",
235
+ "correct": true,
236
+ "predicted": "72",
237
+ "trace_chars": 238
238
+ },
239
+ {
240
+ "id": "m10",
241
+ "correct": true,
242
+ "predicted": "1210",
243
+ "trace_chars": 213
244
+ },
245
+ {
246
+ "id": "m11",
247
+ "correct": true,
248
+ "predicted": "8",
249
+ "trace_chars": 326
250
+ },
251
+ {
252
+ "id": "m12",
253
+ "correct": true,
254
+ "predicted": "28",
255
+ "trace_chars": 341
256
+ },
257
+ {
258
+ "id": "m13",
259
+ "correct": true,
260
+ "predicted": "154",
261
+ "trace_chars": 281
262
+ },
263
+ {
264
+ "id": "m14",
265
+ "correct": true,
266
+ "predicted": "13",
267
+ "trace_chars": 230
268
+ },
269
+ {
270
+ "id": "m15",
271
+ "correct": true,
272
+ "predicted": "75",
273
+ "trace_chars": 261
274
+ },
275
+ {
276
+ "id": "m16",
277
+ "correct": true,
278
+ "predicted": "6",
279
+ "trace_chars": 495
280
+ },
281
+ {
282
+ "id": "m17",
283
+ "correct": true,
284
+ "predicted": "48",
285
+ "trace_chars": 280
286
+ },
287
+ {
288
+ "id": "m18",
289
+ "correct": true,
290
+ "predicted": "18",
291
+ "trace_chars": 340
292
+ },
293
+ {
294
+ "id": "m19",
295
+ "correct": true,
296
+ "predicted": "0.5",
297
+ "trace_chars": 438
298
+ },
299
+ {
300
+ "id": "m20",
301
+ "correct": true,
302
+ "predicted": "216",
303
+ "trace_chars": 165
304
+ },
305
+ {
306
+ "id": "h01",
307
+ "correct": true,
308
+ "predicted": "42",
309
+ "trace_chars": 240
310
+ },
311
+ {
312
+ "id": "h02",
313
+ "correct": true,
314
+ "predicted": "1048576",
315
+ "trace_chars": 203
316
+ },
317
+ {
318
+ "id": "h03",
319
+ "correct": true,
320
+ "predicted": "10",
321
+ "trace_chars": 212
322
+ },
323
+ {
324
+ "id": "h04",
325
+ "correct": true,
326
+ "predicted": "8",
327
+ "trace_chars": 312
328
+ },
329
+ {
330
+ "id": "h05",
331
+ "correct": true,
332
+ "predicted": "100",
333
+ "trace_chars": 271
334
+ },
335
+ {
336
+ "id": "h06",
337
+ "correct": true,
338
+ "predicted": "56",
339
+ "trace_chars": 197
340
+ },
341
+ {
342
+ "id": "h07",
343
+ "correct": true,
344
+ "predicted": "20",
345
+ "trace_chars": 228
346
+ },
347
+ {
348
+ "id": "h08",
349
+ "correct": true,
350
+ "predicted": "10",
351
+ "trace_chars": 147
352
+ },
353
+ {
354
+ "id": "h09",
355
+ "correct": true,
356
+ "predicted": "12",
357
+ "trace_chars": 259
358
+ },
359
+ {
360
+ "id": "h10",
361
+ "correct": true,
362
+ "predicted": "13",
363
+ "trace_chars": 277
364
+ },
365
+ {
366
+ "id": "h11",
367
+ "correct": true,
368
+ "predicted": "17",
369
+ "trace_chars": 121
370
+ },
371
+ {
372
+ "id": "h12",
373
+ "correct": true,
374
+ "predicted": "25",
375
+ "trace_chars": 371
376
+ },
377
+ {
378
+ "id": "h13",
379
+ "correct": true,
380
+ "predicted": "30",
381
+ "trace_chars": 341
382
+ },
383
+ {
384
+ "id": "h14",
385
+ "correct": true,
386
+ "predicted": "132",
387
+ "trace_chars": 180
388
+ },
389
+ {
390
+ "id": "h15",
391
+ "correct": true,
392
+ "predicted": "12",
393
+ "trace_chars": 288
394
+ },
395
+ {
396
+ "id": "h16",
397
+ "correct": true,
398
+ "predicted": "720",
399
+ "trace_chars": 238
400
+ },
401
+ {
402
+ "id": "h17",
403
+ "correct": true,
404
+ "predicted": "1",
405
+ "trace_chars": 328
406
+ },
407
+ {
408
+ "id": "h18",
409
+ "correct": true,
410
+ "predicted": "60",
411
+ "trace_chars": 309
412
+ },
413
+ {
414
+ "id": "h19",
415
+ "correct": true,
416
+ "predicted": "0.375",
417
+ "trace_chars": 381
418
+ },
419
+ {
420
+ "id": "h20",
421
+ "correct": true,
422
+ "predicted": "55",
423
+ "trace_chars": 153
424
+ },
425
+ {
426
+ "id": "i01",
427
+ "correct": true,
428
+ "predicted": "B",
429
+ "trace_chars": 1878
430
+ },
431
+ {
432
+ "id": "i02",
433
+ "correct": false,
434
+ "predicted": null,
435
+ "trace_chars": 1074
436
+ },
437
+ {
438
+ "id": "i03",
439
+ "correct": false,
440
+ "predicted": null,
441
+ "trace_chars": 1413
442
+ },
443
+ {
444
+ "id": "i04",
445
+ "correct": false,
446
+ "predicted": null,
447
+ "trace_chars": 2836
448
+ },
449
+ {
450
+ "id": "i05",
451
+ "correct": false,
452
+ "predicted": null,
453
+ "trace_chars": 1461
454
+ },
455
+ {
456
+ "id": "i06",
457
+ "correct": false,
458
+ "predicted": null,
459
+ "trace_chars": 838
460
+ },
461
+ {
462
+ "id": "i07",
463
+ "correct": false,
464
+ "predicted": null,
465
+ "trace_chars": 1360
466
+ },
467
+ {
468
+ "id": "i08",
469
+ "correct": false,
470
+ "predicted": null,
471
+ "trace_chars": 858
472
+ }
473
+ ]
474
+ },
475
+ "medium": {
476
+ "n": 68,
477
+ "accuracy": 0.8824,
478
+ "median_thinking_tokens": 47.5,
479
+ "empty_trace_rate": 0.0,
480
+ "responses": [
481
+ {
482
+ "id": "e01",
483
+ "correct": true,
484
+ "predicted": "43",
485
+ "trace_chars": 68
486
+ },
487
+ {
488
+ "id": "e02",
489
+ "correct": true,
490
+ "predicted": "62",
491
+ "trace_chars": 159
492
+ },
493
+ {
494
+ "id": "e03",
495
+ "correct": true,
496
+ "predicted": "12",
497
+ "trace_chars": 109
498
+ },
499
+ {
500
+ "id": "e04",
501
+ "correct": true,
502
+ "predicted": "12",
503
+ "trace_chars": 80
504
+ },
505
+ {
506
+ "id": "e05",
507
+ "correct": true,
508
+ "predicted": "150",
509
+ "trace_chars": 88
510
+ },
511
+ {
512
+ "id": "e06",
513
+ "correct": true,
514
+ "predicted": "5.75",
515
+ "trace_chars": 61
516
+ },
517
+ {
518
+ "id": "e07",
519
+ "correct": true,
520
+ "predicted": "33",
521
+ "trace_chars": 108
522
+ },
523
+ {
524
+ "id": "e08",
525
+ "correct": true,
526
+ "predicted": "1024",
527
+ "trace_chars": 192
528
+ },
529
+ {
530
+ "id": "e09",
531
+ "correct": true,
532
+ "predicted": "210",
533
+ "trace_chars": 184
534
+ },
535
+ {
536
+ "id": "e10",
537
+ "correct": true,
538
+ "predicted": "495",
539
+ "trace_chars": 88
540
+ },
541
+ {
542
+ "id": "e11",
543
+ "correct": true,
544
+ "predicted": "28",
545
+ "trace_chars": 182
546
+ },
547
+ {
548
+ "id": "e12",
549
+ "correct": true,
550
+ "predicted": "63",
551
+ "trace_chars": 73
552
+ },
553
+ {
554
+ "id": "e13",
555
+ "correct": true,
556
+ "predicted": "64",
557
+ "trace_chars": 103
558
+ },
559
+ {
560
+ "id": "e14",
561
+ "correct": true,
562
+ "predicted": "144",
563
+ "trace_chars": 76
564
+ },
565
+ {
566
+ "id": "e15",
567
+ "correct": true,
568
+ "predicted": "7",
569
+ "trace_chars": 173
570
+ },
571
+ {
572
+ "id": "e16",
573
+ "correct": true,
574
+ "predicted": "8",
575
+ "trace_chars": 92
576
+ },
577
+ {
578
+ "id": "e17",
579
+ "correct": true,
580
+ "predicted": "12",
581
+ "trace_chars": 145
582
+ },
583
+ {
584
+ "id": "e18",
585
+ "correct": true,
586
+ "predicted": "100",
587
+ "trace_chars": 65
588
+ },
589
+ {
590
+ "id": "e19",
591
+ "correct": true,
592
+ "predicted": "90",
593
+ "trace_chars": 75
594
+ },
595
+ {
596
+ "id": "e20",
597
+ "correct": true,
598
+ "predicted": "125",
599
+ "trace_chars": 74
600
+ },
601
+ {
602
+ "id": "m01",
603
+ "correct": true,
604
+ "predicted": "1081",
605
+ "trace_chars": 158
606
+ },
607
+ {
608
+ "id": "m02",
609
+ "correct": true,
610
+ "predicted": "32768",
611
+ "trace_chars": 225
612
+ },
613
+ {
614
+ "id": "m03",
615
+ "correct": true,
616
+ "predicted": "5050",
617
+ "trace_chars": 176
618
+ },
619
+ {
620
+ "id": "m04",
621
+ "correct": true,
622
+ "predicted": "5",
623
+ "trace_chars": 490
624
+ },
625
+ {
626
+ "id": "m05",
627
+ "correct": true,
628
+ "predicted": "36",
629
+ "trace_chars": 343
630
+ },
631
+ {
632
+ "id": "m06",
633
+ "correct": true,
634
+ "predicted": "12",
635
+ "trace_chars": 236
636
+ },
637
+ {
638
+ "id": "m07",
639
+ "correct": true,
640
+ "predicted": "300",
641
+ "trace_chars": 278
642
+ },
643
+ {
644
+ "id": "m08",
645
+ "correct": true,
646
+ "predicted": "204",
647
+ "trace_chars": 244
648
+ },
649
+ {
650
+ "id": "m09",
651
+ "correct": true,
652
+ "predicted": "72",
653
+ "trace_chars": 245
654
+ },
655
+ {
656
+ "id": "m10",
657
+ "correct": true,
658
+ "predicted": "1210",
659
+ "trace_chars": 220
660
+ },
661
+ {
662
+ "id": "m11",
663
+ "correct": true,
664
+ "predicted": "8",
665
+ "trace_chars": 85
666
+ },
667
+ {
668
+ "id": "m12",
669
+ "correct": true,
670
+ "predicted": "28",
671
+ "trace_chars": 101
672
+ },
673
+ {
674
+ "id": "m13",
675
+ "correct": true,
676
+ "predicted": "154",
677
+ "trace_chars": 211
678
+ },
679
+ {
680
+ "id": "m14",
681
+ "correct": true,
682
+ "predicted": "13",
683
+ "trace_chars": 187
684
+ },
685
+ {
686
+ "id": "m15",
687
+ "correct": true,
688
+ "predicted": "75",
689
+ "trace_chars": 105
690
+ },
691
+ {
692
+ "id": "m16",
693
+ "correct": true,
694
+ "predicted": "6",
695
+ "trace_chars": 219
696
+ },
697
+ {
698
+ "id": "m17",
699
+ "correct": true,
700
+ "predicted": "48",
701
+ "trace_chars": 158
702
+ },
703
+ {
704
+ "id": "m18",
705
+ "correct": true,
706
+ "predicted": "18",
707
+ "trace_chars": 220
708
+ },
709
+ {
710
+ "id": "m19",
711
+ "correct": true,
712
+ "predicted": "0.5",
713
+ "trace_chars": 245
714
+ },
715
+ {
716
+ "id": "m20",
717
+ "correct": true,
718
+ "predicted": "216",
719
+ "trace_chars": 101
720
+ },
721
+ {
722
+ "id": "h01",
723
+ "correct": true,
724
+ "predicted": "42",
725
+ "trace_chars": 271
726
+ },
727
+ {
728
+ "id": "h02",
729
+ "correct": true,
730
+ "predicted": "1048576",
731
+ "trace_chars": 299
732
+ },
733
+ {
734
+ "id": "h03",
735
+ "correct": true,
736
+ "predicted": "10",
737
+ "trace_chars": 184
738
+ },
739
+ {
740
+ "id": "h04",
741
+ "correct": true,
742
+ "predicted": "8",
743
+ "trace_chars": 136
744
+ },
745
+ {
746
+ "id": "h05",
747
+ "correct": true,
748
+ "predicted": "100",
749
+ "trace_chars": 341
750
+ },
751
+ {
752
+ "id": "h06",
753
+ "correct": true,
754
+ "predicted": "56",
755
+ "trace_chars": 197
756
+ },
757
+ {
758
+ "id": "h07",
759
+ "correct": true,
760
+ "predicted": "20",
761
+ "trace_chars": 198
762
+ },
763
+ {
764
+ "id": "h08",
765
+ "correct": true,
766
+ "predicted": "10",
767
+ "trace_chars": 183
768
+ },
769
+ {
770
+ "id": "h09",
771
+ "correct": true,
772
+ "predicted": "12",
773
+ "trace_chars": 208
774
+ },
775
+ {
776
+ "id": "h10",
777
+ "correct": true,
778
+ "predicted": "13",
779
+ "trace_chars": 119
780
+ },
781
+ {
782
+ "id": "h11",
783
+ "correct": true,
784
+ "predicted": "17",
785
+ "trace_chars": 165
786
+ },
787
+ {
788
+ "id": "h12",
789
+ "correct": true,
790
+ "predicted": "25",
791
+ "trace_chars": 176
792
+ },
793
+ {
794
+ "id": "h13",
795
+ "correct": true,
796
+ "predicted": "30",
797
+ "trace_chars": 269
798
+ },
799
+ {
800
+ "id": "h14",
801
+ "correct": true,
802
+ "predicted": "132",
803
+ "trace_chars": 164
804
+ },
805
+ {
806
+ "id": "h15",
807
+ "correct": true,
808
+ "predicted": "12",
809
+ "trace_chars": 447
810
+ },
811
+ {
812
+ "id": "h16",
813
+ "correct": true,
814
+ "predicted": "720",
815
+ "trace_chars": 235
816
+ },
817
+ {
818
+ "id": "h17",
819
+ "correct": true,
820
+ "predicted": "1",
821
+ "trace_chars": 406
822
+ },
823
+ {
824
+ "id": "h18",
825
+ "correct": true,
826
+ "predicted": "60",
827
+ "trace_chars": 460
828
+ },
829
+ {
830
+ "id": "h19",
831
+ "correct": true,
832
+ "predicted": "0.375",
833
+ "trace_chars": 198
834
+ },
835
+ {
836
+ "id": "h20",
837
+ "correct": true,
838
+ "predicted": "55",
839
+ "trace_chars": 196
840
+ },
841
+ {
842
+ "id": "i01",
843
+ "correct": false,
844
+ "predicted": "A",
845
+ "trace_chars": 1944
846
+ },
847
+ {
848
+ "id": "i02",
849
+ "correct": false,
850
+ "predicted": "A",
851
+ "trace_chars": 1057
852
+ },
853
+ {
854
+ "id": "i03",
855
+ "correct": false,
856
+ "predicted": "B",
857
+ "trace_chars": 1272
858
+ },
859
+ {
860
+ "id": "i04",
861
+ "correct": false,
862
+ "predicted": "A",
863
+ "trace_chars": 967
864
+ },
865
+ {
866
+ "id": "i05",
867
+ "correct": false,
868
+ "predicted": "A",
869
+ "trace_chars": 1825
870
+ },
871
+ {
872
+ "id": "i06",
873
+ "correct": false,
874
+ "predicted": "B",
875
+ "trace_chars": 1222
876
+ },
877
+ {
878
+ "id": "i07",
879
+ "correct": false,
880
+ "predicted": "A",
881
+ "trace_chars": 1738
882
+ },
883
+ {
884
+ "id": "i08",
885
+ "correct": false,
886
+ "predicted": "A",
887
+ "trace_chars": 1218
888
+ }
889
+ ]
890
+ },
891
+ "low": {
892
+ "n": 68,
893
+ "accuracy": 0.8971,
894
+ "median_thinking_tokens": 45.0,
895
+ "empty_trace_rate": 0.0,
896
+ "responses": [
897
+ {
898
+ "id": "e01",
899
+ "correct": true,
900
+ "predicted": "43",
901
+ "trace_chars": 51
902
+ },
903
+ {
904
+ "id": "e02",
905
+ "correct": true,
906
+ "predicted": "62",
907
+ "trace_chars": 158
908
+ },
909
+ {
910
+ "id": "e03",
911
+ "correct": true,
912
+ "predicted": "12",
913
+ "trace_chars": 109
914
+ },
915
+ {
916
+ "id": "e04",
917
+ "correct": true,
918
+ "predicted": "12",
919
+ "trace_chars": 61
920
+ },
921
+ {
922
+ "id": "e05",
923
+ "correct": true,
924
+ "predicted": "150",
925
+ "trace_chars": 110
926
+ },
927
+ {
928
+ "id": "e06",
929
+ "correct": true,
930
+ "predicted": "5.75",
931
+ "trace_chars": 61
932
+ },
933
+ {
934
+ "id": "e07",
935
+ "correct": true,
936
+ "predicted": "33",
937
+ "trace_chars": 76
938
+ },
939
+ {
940
+ "id": "e08",
941
+ "correct": true,
942
+ "predicted": "1024",
943
+ "trace_chars": 192
944
+ },
945
+ {
946
+ "id": "e09",
947
+ "correct": true,
948
+ "predicted": "210",
949
+ "trace_chars": 170
950
+ },
951
+ {
952
+ "id": "e10",
953
+ "correct": true,
954
+ "predicted": "495",
955
+ "trace_chars": 88
956
+ },
957
+ {
958
+ "id": "e11",
959
+ "correct": true,
960
+ "predicted": "28",
961
+ "trace_chars": 130
962
+ },
963
+ {
964
+ "id": "e12",
965
+ "correct": true,
966
+ "predicted": "63",
967
+ "trace_chars": 73
968
+ },
969
+ {
970
+ "id": "e13",
971
+ "correct": true,
972
+ "predicted": "64",
973
+ "trace_chars": 80
974
+ },
975
+ {
976
+ "id": "e14",
977
+ "correct": true,
978
+ "predicted": "144",
979
+ "trace_chars": 58
980
+ },
981
+ {
982
+ "id": "e15",
983
+ "correct": true,
984
+ "predicted": "7",
985
+ "trace_chars": 139
986
+ },
987
+ {
988
+ "id": "e16",
989
+ "correct": true,
990
+ "predicted": "8",
991
+ "trace_chars": 84
992
+ },
993
+ {
994
+ "id": "e17",
995
+ "correct": true,
996
+ "predicted": "12",
997
+ "trace_chars": 173
998
+ },
999
+ {
1000
+ "id": "e18",
1001
+ "correct": true,
1002
+ "predicted": "100",
1003
+ "trace_chars": 65
1004
+ },
1005
+ {
1006
+ "id": "e19",
1007
+ "correct": true,
1008
+ "predicted": "90",
1009
+ "trace_chars": 73
1010
+ },
1011
+ {
1012
+ "id": "e20",
1013
+ "correct": true,
1014
+ "predicted": "125",
1015
+ "trace_chars": 74
1016
+ },
1017
+ {
1018
+ "id": "m01",
1019
+ "correct": true,
1020
+ "predicted": "1081",
1021
+ "trace_chars": 91
1022
+ },
1023
+ {
1024
+ "id": "m02",
1025
+ "correct": true,
1026
+ "predicted": "32768",
1027
+ "trace_chars": 235
1028
+ },
1029
+ {
1030
+ "id": "m03",
1031
+ "correct": true,
1032
+ "predicted": "5050",
1033
+ "trace_chars": 220
1034
+ },
1035
+ {
1036
+ "id": "m04",
1037
+ "correct": true,
1038
+ "predicted": "5",
1039
+ "trace_chars": 528
1040
+ },
1041
+ {
1042
+ "id": "m05",
1043
+ "correct": true,
1044
+ "predicted": "36",
1045
+ "trace_chars": 280
1046
+ },
1047
+ {
1048
+ "id": "m06",
1049
+ "correct": true,
1050
+ "predicted": "12",
1051
+ "trace_chars": 221
1052
+ },
1053
+ {
1054
+ "id": "m07",
1055
+ "correct": true,
1056
+ "predicted": "300",
1057
+ "trace_chars": 241
1058
+ },
1059
+ {
1060
+ "id": "m08",
1061
+ "correct": true,
1062
+ "predicted": "204",
1063
+ "trace_chars": 246
1064
+ },
1065
+ {
1066
+ "id": "m09",
1067
+ "correct": true,
1068
+ "predicted": "72",
1069
+ "trace_chars": 198
1070
+ },
1071
+ {
1072
+ "id": "m10",
1073
+ "correct": true,
1074
+ "predicted": "1210",
1075
+ "trace_chars": 200
1076
+ },
1077
+ {
1078
+ "id": "m11",
1079
+ "correct": true,
1080
+ "predicted": "8",
1081
+ "trace_chars": 155
1082
+ },
1083
+ {
1084
+ "id": "m12",
1085
+ "correct": true,
1086
+ "predicted": "28",
1087
+ "trace_chars": 152
1088
+ },
1089
+ {
1090
+ "id": "m13",
1091
+ "correct": true,
1092
+ "predicted": "154",
1093
+ "trace_chars": 144
1094
+ },
1095
+ {
1096
+ "id": "m14",
1097
+ "correct": true,
1098
+ "predicted": "13",
1099
+ "trace_chars": 162
1100
+ },
1101
+ {
1102
+ "id": "m15",
1103
+ "correct": true,
1104
+ "predicted": "75",
1105
+ "trace_chars": 90
1106
+ },
1107
+ {
1108
+ "id": "m16",
1109
+ "correct": true,
1110
+ "predicted": "6",
1111
+ "trace_chars": 180
1112
+ },
1113
+ {
1114
+ "id": "m17",
1115
+ "correct": true,
1116
+ "predicted": "48",
1117
+ "trace_chars": 160
1118
+ },
1119
+ {
1120
+ "id": "m18",
1121
+ "correct": true,
1122
+ "predicted": "18",
1123
+ "trace_chars": 220
1124
+ },
1125
+ {
1126
+ "id": "m19",
1127
+ "correct": true,
1128
+ "predicted": "0.5",
1129
+ "trace_chars": 123
1130
+ },
1131
+ {
1132
+ "id": "m20",
1133
+ "correct": true,
1134
+ "predicted": "216",
1135
+ "trace_chars": 115
1136
+ },
1137
+ {
1138
+ "id": "h01",
1139
+ "correct": true,
1140
+ "predicted": "42",
1141
+ "trace_chars": 252
1142
+ },
1143
+ {
1144
+ "id": "h02",
1145
+ "correct": true,
1146
+ "predicted": "1048576",
1147
+ "trace_chars": 299
1148
+ },
1149
+ {
1150
+ "id": "h03",
1151
+ "correct": true,
1152
+ "predicted": "10",
1153
+ "trace_chars": 121
1154
+ },
1155
+ {
1156
+ "id": "h04",
1157
+ "correct": true,
1158
+ "predicted": "8",
1159
+ "trace_chars": 566
1160
+ },
1161
+ {
1162
+ "id": "h05",
1163
+ "correct": true,
1164
+ "predicted": "100",
1165
+ "trace_chars": 275
1166
+ },
1167
+ {
1168
+ "id": "h06",
1169
+ "correct": true,
1170
+ "predicted": "56",
1171
+ "trace_chars": 196
1172
+ },
1173
+ {
1174
+ "id": "h07",
1175
+ "correct": true,
1176
+ "predicted": "20",
1177
+ "trace_chars": 179
1178
+ },
1179
+ {
1180
+ "id": "h08",
1181
+ "correct": true,
1182
+ "predicted": "10",
1183
+ "trace_chars": 187
1184
+ },
1185
+ {
1186
+ "id": "h09",
1187
+ "correct": true,
1188
+ "predicted": "12",
1189
+ "trace_chars": 206
1190
+ },
1191
+ {
1192
+ "id": "h10",
1193
+ "correct": true,
1194
+ "predicted": "13",
1195
+ "trace_chars": 138
1196
+ },
1197
+ {
1198
+ "id": "h11",
1199
+ "correct": true,
1200
+ "predicted": "17",
1201
+ "trace_chars": 156
1202
+ },
1203
+ {
1204
+ "id": "h12",
1205
+ "correct": true,
1206
+ "predicted": "25",
1207
+ "trace_chars": 157
1208
+ },
1209
+ {
1210
+ "id": "h13",
1211
+ "correct": true,
1212
+ "predicted": "30",
1213
+ "trace_chars": 291
1214
+ },
1215
+ {
1216
+ "id": "h14",
1217
+ "correct": true,
1218
+ "predicted": "132",
1219
+ "trace_chars": 161
1220
+ },
1221
+ {
1222
+ "id": "h15",
1223
+ "correct": true,
1224
+ "predicted": "12",
1225
+ "trace_chars": 404
1226
+ },
1227
+ {
1228
+ "id": "h16",
1229
+ "correct": true,
1230
+ "predicted": "720",
1231
+ "trace_chars": 216
1232
+ },
1233
+ {
1234
+ "id": "h17",
1235
+ "correct": true,
1236
+ "predicted": "1",
1237
+ "trace_chars": 267
1238
+ },
1239
+ {
1240
+ "id": "h18",
1241
+ "correct": true,
1242
+ "predicted": "60",
1243
+ "trace_chars": 464
1244
+ },
1245
+ {
1246
+ "id": "h19",
1247
+ "correct": true,
1248
+ "predicted": "0.375",
1249
+ "trace_chars": 288
1250
+ },
1251
+ {
1252
+ "id": "h20",
1253
+ "correct": true,
1254
+ "predicted": "55",
1255
+ "trace_chars": 196
1256
+ },
1257
+ {
1258
+ "id": "i01",
1259
+ "correct": false,
1260
+ "predicted": "A",
1261
+ "trace_chars": 1745
1262
+ },
1263
+ {
1264
+ "id": "i02",
1265
+ "correct": false,
1266
+ "predicted": "A",
1267
+ "trace_chars": 949
1268
+ },
1269
+ {
1270
+ "id": "i03",
1271
+ "correct": false,
1272
+ "predicted": "B",
1273
+ "trace_chars": 1123
1274
+ },
1275
+ {
1276
+ "id": "i04",
1277
+ "correct": false,
1278
+ "predicted": "A",
1279
+ "trace_chars": 917
1280
+ },
1281
+ {
1282
+ "id": "i05",
1283
+ "correct": false,
1284
+ "predicted": "A",
1285
+ "trace_chars": 1428
1286
+ },
1287
+ {
1288
+ "id": "i06",
1289
+ "correct": true,
1290
+ "predicted": "A",
1291
+ "trace_chars": 1034
1292
+ },
1293
+ {
1294
+ "id": "i07",
1295
+ "correct": false,
1296
+ "predicted": "B",
1297
+ "trace_chars": 1593
1298
+ },
1299
+ {
1300
+ "id": "i08",
1301
+ "correct": false,
1302
+ "predicted": "A",
1303
+ "trace_chars": 1314
1304
+ }
1305
+ ]
1306
+ }
1307
+ }
1308
+ }