{
  "schema_version": 1,
  "title": "Repro - Maximum Likelihood Reinforcement Learning",
  "emoji": "\ud83d\udd2c",
  "space_id": "bertfil/EeuLO2BjFN",
  "paper": {
    "arxiv_id": "2602.02710",
    "openreview_id": "EeuLO2BjFN",
    "title": "Maximum Likelihood Reinforcement Learning",
    "url": "https://arxiv.org/abs/2602.02710"
  },
  "tags": [
    "icml2026-repro",
    "paper-EeuLO2BjFN"
  ],
  "updated_at": "2026-07-20T03:52:35+00:00",
  "root": {
    "slug": "index",
    "title": "Repro - Maximum Likelihood Reinforcement Learning",
    "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-maxrl-defines-a-compute-indexed-family-of-sample-b",
        "title": "Claim 1 - MaxRL defines a compute-indexed family of sample-based objec\u2026",
        "file": "pages/claim-1-maxrl-defines-a-compute-indexed-family-of-sample-b/page.md",
        "children": []
      },
      {
        "slug": "claim-2-the-maxrl-objectives-admit-a-simple-unbiased-polic",
        "title": "Claim 2 - The MaxRL objectives admit a simple unbiased policy-gradient\u2026",
        "file": "pages/claim-2-the-maxrl-objectives-admit-a-simple-unbiased-polic/page.md",
        "children": []
      },
      {
        "slug": "claim-3-the-paper-claims-maxrl-converges-to-maximum-likeli",
        "title": "Claim 3 - The paper claims MaxRL converges to maximum-likelihood optim\u2026",
        "file": "pages/claim-3-the-paper-claims-maxrl-converges-to-maximum-likeli/page.md",
        "children": []
      },
      {
        "slug": "claim-4-empirically-maxrl-pareto-dominates-tested-existing",
        "title": "Claim 4 - Empirically, MaxRL Pareto-dominates tested existing methods \u2026",
        "file": "pages/claim-4-empirically-maxrl-pareto-dominates-tested-existing/page.md",
        "children": []
      },
      {
        "slug": "claim-5-maxrl-reports-up-to-20x-test-time-scaling-efficien",
        "title": "Claim 5 - MaxRL reports up to 20x test-time scaling efficiency gains c\u2026",
        "file": "pages/claim-5-maxrl-reports-up-to-20x-test-time-scaling-efficien/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": 8511,
  "revision": "1784519555880070100"
}