Tadrisino/MathApp
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

keep the raw benchmark answers so the report's questions and answers never need re-running

Browse Source 
102 unique textbook questions, 1,287 full model answers, per-answer cost and
the hand grades for both rounds. Book excerpts (retrieval context) are left out.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
Zareian
2026-09-26 10:43:43 +03:30
parent 64cf018e55
commit 5d490edaf9
19 changed files with 25668 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

228
docs/benchmarks/2026-09-26-tutor-model/round2.json Normal file
View File

@@ -0,0 +1,228 @@
{
"v3-4o": {
"total": 86,
"n": 102,
"tier": {
"primary": [
38,
46
],
"middle": [
22,
26
],
"upper": [
26,
30
]
},
"slash": [
13,
17
],
"cents": 0.0410316990291262,
"p50": 5.9,
"markup": 0.019417475728155338,
"wrong": [
313,
422,
511,
519,
521,
592,
998,
1051,
1398,
2534,
2771,
2839,
3591,
3935,
3959,
4113
]
},
"v4-4o": {
"total": 90,
"n": 102,
"tier": {
"primary": [
38,
46
],
"middle": [
25,
26
],
"upper": [
27,
30
]
},
"slash": [
13,
17
],
"cents": 0.041762038834951457,
"p50": 5.2,
"markup": 0.0,
"wrong": [
313,
422,
519,
521,
592,
621,
1051,
1398,
2534,
2839,
3935,
4113
]
},
"v5-4o": {
"total": 85,
"n": 102,
"tier": {
"primary": [
38,
46
],
"middle": [
22,
26
],
"upper": [
25,
30
]
},
"slash": [
12,
17
],
"cents": 0.04241577669902913,
"p50": 3.9,
"markup": 0.009708737864077669,
"wrong": [
312,
313,
393,
422,
519,
521,
592,
907,
998,
1398,
2534,
2839,
2930,
3935,
3959,
4108,
4113
]
},
"v3-41": {
"total": 95,
"n": 102,
"tier": {
"primary": [
39,
46
],
"middle": [
26,
26
],
"upper": [
30,
30
]
},
"slash": [
11,
17
],
"cents": 0.11057048543689321,
"p50": 4.0,
"markup": 0.0,
"wrong": [
313,
393,
397,
422,
511,
621,
626
]
},
"v4-41": {
"total": 96,
"n": 102,
"tier": {
"primary": [
40,
46
],
"middle": [
26,
26
],
"upper": [
30,
30
]
},
"slash": [
11,
17
],
"cents": 0.11649165048543685,
"p50": 3.5,
"markup": 0.0,
"wrong": [
313,
393,
397,
511,
621,
626
]
},
"v5-41": {
"total": 96,
"n": 102,
"tier": {
"primary": [
42,
46
],
"middle": [
25,
26
],
"upper": [
29,
30
]
},
"slash": [
14,
17
],
"cents": 0.12201592233009707,
"p50": 3.3,
"markup": 0.0,
"wrong": [
313,
397,
422,
511,
2839,
3938
]
}
}