File size: 3,257 Bytes
6e3a665
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9fd3976
6e3a665
 
 
 
 
 
 
 
 
 
 
 
9fd3976
 
 
6e3a665
 
 
9fd3976
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
6e3a665
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9fd3976
 
6e3a665
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
{
  "schema_version": 1,
  "title": "Repro - Prescriptive Scaling Reveals the Evolution of Language Model Capabilities",
  "emoji": "\ud83d\udd2c",
  "space_id": "bertfil/IkjsHRpuYY",
  "paper": {
    "arxiv_id": "2602.15327",
    "openreview_id": "IkjsHRpuYY",
    "title": "Prescriptive Scaling Reveals the Evolution of Language Model Capabilities",
    "url": "https://arxiv.org/abs/2602.15327"
  },
  "tags": [
    "icml2026-repro",
    "paper-IkjsHRpuYY"
  ],
  "updated_at": "2026-07-20T03:52:38+00:00",
  "root": {
    "slug": "index",
    "title": "Repro - Prescriptive Scaling Reveals the Evolution of Language Model Capabilities",
    "file": "pages/index.md",
    "children": [
      {
        "slug": "00-scored-evidence-summary",
        "title": "00 - Scored evidence summary",
        "file": "pages/00-scored-evidence-summary/page.md",
        "children": []
      },
      {
        "slug": "claim-1-capability-boundaries-are-estimated-via-smoothed-p",
        "title": "Claim 1 - Capability boundaries are estimated via smoothed pinball-los\u2026",
        "file": "pages/claim-1-capability-boundaries-are-estimated-via-smoothed-p/page.md",
        "children": []
      },
      {
        "slug": "claim-2-post-training-capability-boundaries-are-modeled-as",
        "title": "Claim 2 - Post-training capability boundaries are modeled as sigmoid f\u2026",
        "file": "pages/claim-2-post-training-capability-boundaries-are-modeled-as/page.md",
        "children": []
      },
      {
        "slug": "claim-3-at-10-flops-the-estimated-0-98-quantile-boundary-r",
        "title": "Claim 3 - At 10\u00b2\u2074 FLOPs, the estimated 0.98-quantile boundary reaches \u2026",
        "file": "pages/claim-3-at-10-flops-the-estimated-0-98-quantile-boundary-r/page.md",
        "children": []
      },
      {
        "slug": "claim-4-unlike-the-stable-boundaries-for-bbh-gpqa-mmlu-pro",
        "title": "Claim 4 - Unlike the stable boundaries for BBH, GPQA, MMLU-Pro, and MU\u2026",
        "file": "pages/claim-4-unlike-the-stable-boundaries-for-bbh-gpqa-mmlu-pro/page.md",
        "children": []
      },
      {
        "slug": "claim-5-a-balanced-i-optimal-sampling-design-recovers-near",
        "title": "Claim 5 - A balanced I-optimal sampling design recovers near-full-data\u2026",
        "file": "pages/claim-5-a-balanced-i-optimal-sampling-design-recovers-near/page.md",
        "children": []
      },
      {
        "slug": "claim-6-a-cross-benchmark-shift-test-applied-to-aime-2025",
        "title": "Claim 6 - A cross-benchmark shift test applied to AIME-2025 scores fin\u2026",
        "file": "pages/claim-6-a-cross-benchmark-shift-test-applied-to-aime-2025/page.md",
        "children": []
      },
      {
        "slug": "reproduction-protocol-and-provenance",
        "title": "Reproduction protocol and provenance",
        "file": "pages/reproduction-protocol-and-provenance/page.md",
        "children": []
      },
      {
        "slug": "conclusion",
        "title": "Conclusion",
        "file": "pages/conclusion/page.md",
        "children": []
      }
    ]
  },
  "agent_view_tokens": 6563,
  "revision": "1784519558058211500"
}