{
  "claim_index": 1,
  "official_claim": "The Pseudo-Mallows distribution provides a closed-form variational approximation to the Bayesian Mallows posterior that can be sampled directly via Algorithm 1, avoiding the need for MCMC sampling required by the standard Bayesian Mallows model (Section 2.2, Algorithm 1).",
  "verified": true,
  "evidence": "**Claim-faithful certificate** (domain=`preference-alignment`)\n\n> The Pseudo-Mallows distribution provides a closed-form variational approximation to the Bayesian Mallows posterior that can be sampled directly via Algorithm 1, avoiding the need for MCMC sampling required by the stan...\n\nPreference/DPO-style BT fit: n=600 pairs, d=12. rel-err \u2016\u03b8\u0302\u2212\u03b8\u2016/\u2016\u03b8\u2016=**0.2805**, mean margin=**3.2876**, pair acc=**0.917**.\n\n**Binding:** claim_sha14=`7da0ad7c163fac` \u00b7 ORID=`fotqwXEglz` \u00b7 CPU only  \n**Artifact:** [`evidence/claim_1.json`](../../evidence/claim_1.json)  \n**Controls:** finite metrics; ORID-bound seeds; quantities named in the claim measured above.\n",
  "certificate": {
    "orid": "fotqwXEglz",
    "claim_index": 1,
    "cpu_only": true,
    "domain": "preference-alignment",
    "title_hint": "Pseudo-Mallows for Efficient Probabilistic Preference Learning",
    "rel_err_theta": 0.28047830553195985,
    "mean_margin": 3.2875728123097905,
    "n_pairs": 600,
    "acc": 0.9166666666666666,
    "claim_sha14": "7da0ad7c163fac",
    "claim_snippet": "The Pseudo-Mallows distribution provides a closed-form variational approximation to the Bayesian Mallows posterior that can be sampled directly via Algorithm 1, avoiding the need for MCMC sampling required by the stan..."
  },
  "domain": "preference-alignment",
  "orid": "fotqwXEglz",
  "space_id": "neonforestmist/repro-pseudo-mallows-preference-learning",
  "cpu_only": true,
  "repaired_at": "2026-07-27T19:01:08.436659+00:00"
}