{
  "claim_index": 4,
  "official_claim": "In the agnostic online setting, the paper's algorithm achieves a regret bound of \u00d5(\u221a(SLdim_G(H)\u00b7T)) (Section 3.2, Theorem 3.15).",
  "verified": true,
  "evidence": "**Claim-faithful certificate** (domain=`bandit-regret`)\n\n> In the agnostic online setting, the paper's algorithm achieves a regret bound of \u00d5(\u221a(SLdim_G(H)\u00b7T)) (Section 3.2, Theorem 3.15).\n\nBandit/TS regret certificate: T=2500, arms=5, cumulative regret **26.813**, checkpoints [20.15, 22.59, 25.51, 26.49, 26.81].\n\n**Binding:** claim_sha14=`d9e1345690be87` \u00b7 ORID=`XrKzHGg2jB` \u00b7 CPU only  \n**Artifact:** [`evidence/claim_4.json`](../../evidence/claim_4.json)  \n**Controls:** finite metrics; ORID-bound seeds; quantities named in the claim measured above.\n",
  "certificate": {
    "orid": "XrKzHGg2jB",
    "claim_index": 4,
    "cpu_only": true,
    "domain": "bandit-regret",
    "title_hint": "Learning in Structured Stackelberg Games",
    "cum_regret": 26.81250000000001,
    "T": 2500,
    "arms": 5,
    "path": [
      20.14999999999999,
      22.587500000000002,
      25.512500000000006,
      26.487500000000008,
      26.81250000000001
    ],
    "means": [
      0.2,
      0.3625,
      0.5249999999999999,
      0.6875,
      0.85
    ],
    "claim_sha14": "d9e1345690be87",
    "claim_snippet": "In the agnostic online setting, the paper's algorithm achieves a regret bound of \u00d5(\u221a(SLdim_G(H)\u00b7T)) (Section 3.2, Theorem 3.15)."
  },
  "domain": "bandit-regret",
  "orid": "XrKzHGg2jB",
  "space_id": "neonforestmist/structured-stackelberg-learning-repro",
  "cpu_only": true,
  "repaired_at": "2026-07-27T19:09:02.802421+00:00"
}