File size: 3,550 Bytes
589b85d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
81
82
83
84
85
86
87
{
  "schema_version": 2,
  "title": "Reproduction: MASH: Modeling Abstention via Selective Help-Seeking",
  "emoji": "🎯",
  "space_id": "SabaPivot/repro-mash-selective-help-seeking",
  "paper": {
    "arxiv_id": "2510.01152"
  },
  "tags": [
    "icml2026-repro",
    "paper-d5rd1zlHTf"
  ],
  "updated_at": "2026-07-28T14:49:50+00:00",
  "root": {
    "slug": "index",
    "title": "Reproduction: MASH: Modeling Abstention via Selective Help-Seeking",
    "file": "pages/index.md",
    "children": [
      {
        "slug": "executive-summary",
        "title": "Executive summary",
        "file": "pages/executive-summary/page.md",
        "children": []
      },
      {
        "slug": "claim-1-on-hotpotqa-mash-improves-answer-accuracy-by-7-6-percentage-points",
        "title": "Claim 1: On HotPotQA, MASH improves answer accuracy by 7.6 percentage points",
        "file": "pages/claim-1-on-hotpotqa-mash-improves-answer-accuracy-by-7-6-percentage-points/page.md",
        "children": []
      },
      {
        "slug": "claim-2-on-2wikimultihopqa-mash-improves-accuracy-by-6-64-percentage-points-over",
        "title": "Claim 2: On 2WikiMultihopQA, MASH improves accuracy by 6.64 percentage points over",
        "file": "pages/claim-2-on-2wikimultihopqa-mash-improves-accuracy-by-6-64-percentage-points-over/page.md",
        "children": []
      },
      {
        "slug": "claim-3-mash-improves-average-tool-productivity-by-3-83-points-across-evaluated",
        "title": "Claim 3: MASH improves average Tool Productivity by 3.83 points across evaluated",
        "file": "pages/claim-3-mash-improves-average-tool-productivity-by-3-83-points-across-evaluated/page.md",
        "children": []
      },
      {
        "slug": "claim-4-on-the-subset-of-hotpotqa-questions-requiring-exactly-one-search",
        "title": "Claim 4: On the subset of HotPotQA questions requiring exactly one search",
        "file": "pages/claim-4-on-the-subset-of-hotpotqa-questions-requiring-exactly-one-search/page.md",
        "children": []
      },
      {
        "slug": "claim-5-in-the-abstention-inference-mode-mash-with-otc-strict-reaches-59-9",
        "title": "Claim 5: In the abstention inference mode, MASH with OTC-Strict reaches 59.9%",
        "file": "pages/claim-5-in-the-abstention-inference-mode-mash-with-otc-strict-reaches-59-9/page.md",
        "children": []
      },
      {
        "slug": "claim-6-mash-is-trained-via-a-two-step-pipeline-of-warm-start-supervised",
        "title": "Claim 6: MASH is trained via a two-step pipeline of warm-start supervised",
        "file": "pages/claim-6-mash-is-trained-via-a-two-step-pipeline-of-warm-start-supervised/page.md",
        "children": []
      },
      {
        "slug": "conclusion",
        "title": "Conclusion",
        "file": "pages/conclusion/page.md",
        "children": []
      }
    ]
  },
  "traces": [],
  "workspace": {
    "file": "workspace.json",
    "file_count": 1,
    "total_size": 23331,
    "bucket_id": "SabaPivot/repro-mash-selective-help-seeking-artifacts"
  },
  "agent_view_tokens": 3800,
  "trace_view_tokens": 10,
  "workspace_view_tokens": 18,
  "revision": "b6ce1f04feac8570c199",
  "workspace_ref": {
    "repo_id": "SabaPivot/repro-mash-selective-help-seeking-artifacts",
    "repo_type": "bucket",
    "repo_url": "https://huggingface.co/buckets/SabaPivot/repro-mash-selective-help-seeking-artifacts",
    "private": false
  },
  "workspace_bucket": "https://huggingface.co/buckets/SabaPivot/repro-mash-selective-help-seeking-artifacts"
}