P3Mnh7mF5a / logbook.json
DineshAI's picture
Update logbook: Repro - Provable Benefits of RLVR over SFT for Reasoning Models
538a3cb verified
Raw
History Blame Contribute Delete
1.93 kB
{
"schema_version": 1,
"title": "Repro - Provable Benefits of RLVR over SFT for Reasoning Models",
"emoji": "🎯",
"space_id": "DineshAI/P3Mnh7mF5a",
"paper": null,
"tags": [
"icml2026-repro",
"paper-P3Mnh7mF5a"
],
"updated_at": "2026-07-30T09:52:03+00:00",
"root": {
"slug": "index",
"title": "Repro - Provable Benefits of RLVR over SFT for Reasoning Models",
"file": "pages/index.md",
"children": [
{
"slug": "claim-1-golden-path-sft",
"title": "Claim 1 β€” Golden-path SFT",
"file": "pages/claim-1-golden-path-sft/page.md",
"children": []
},
{
"slug": "claim-2-rlvr-backtracking",
"title": "Claim 2 β€” RLVR backtracking",
"file": "pages/claim-2-rlvr-backtracking/page.md",
"children": []
},
{
"slug": "claim-3-inference-separation",
"title": "Claim 3 β€” Inference separation",
"file": "pages/claim-3-inference-separation/page.md",
"children": []
},
{
"slug": "claim-4-search-agent",
"title": "Claim 4 β€” Search agent",
"file": "pages/claim-4-search-agent/page.md",
"children": []
},
{
"slug": "claim-5-trace-distillation",
"title": "Claim 5 β€” Trace distillation",
"file": "pages/claim-5-trace-distillation/page.md",
"children": []
},
{
"slug": "methods",
"title": "Methods",
"file": "pages/methods/page.md",
"children": []
},
{
"slug": "negative-controls",
"title": "Negative controls",
"file": "pages/negative-controls/page.md",
"children": []
},
{
"slug": "conclusion",
"title": "Conclusion",
"file": "pages/conclusion/page.md",
"children": []
}
]
},
"agent_view_tokens": 1629,
"revision": "1785405123517877047"
}