{
  "asOf": "2026-09-21",
  "leaderboardCheckedAt": "2026-09-22T01:15:34Z",
  "officialBest": 0.328,
  "experiments": [
    {
      "id": "control",
      "type": "official",
      "title": "Reproduce the public baseline",
      "subtitle": "The current scored control · September 21",
      "score": "0.240 → 0.328",
      "verdict": "Adopted control",
      "tone": "",
      "finding": "Official submission 56443775 scored 0.328. The cached control preserved all 17 frozen scoring cells. Its recorded cloud runtime fell from 3,646 to 2,259 seconds; all 400 placeholder ranked lists and CSV bytes matched the slower control.",
      "meaning": "This is an official improvement for the full recipe, not a proven effect of one channel. The predictive design and weights are a public-baseline reproduction. Caching is an execution improvement, not an accuracy gain."
    },
    {
      "id": "transplant",
      "type": "official",
      "title": "Expand the direct-match library",
      "subtitle": "A local prediction did not transfer · September 21",
      "score": "0.240 → 0.238",
      "verdict": "Not promoted",
      "tone": "reject",
      "finding": "The full-library transplant cleared its local gate with predicted leaderboard delta +0.012120 [0.009206, 0.015024]. Official submission 56442933 instead scored 0.238 against the earlier 0.240 comparator.",
      "meaning": "The official test did not confirm the expected gain. One noisy submission does not establish a true 0.002 regression, but this change has no demonstrated benefit."
    },
    {
      "id": "ensemble",
      "type": "local",
      "title": "Combine four neural models",
      "subtitle": "Historical held-out C2 · same candidates",
      "score": "+0.020272",
      "verdict": "Uncertain gain",
      "tone": "hold",
      "finding": "C2 MRR moved from 0.562766 to 0.583038. The paired 95% interval for the change was [−0.004989, +0.045697], crossing zero. Candidate lists and all 43,230 classical-feature rows matched.",
      "meaning": "The primary confidence-interval gate failed. Aligned ranker training and independent, identity-audited evaluation are needed before promotion. This is not a leaderboard score."
    },
    {
      "id": "retrieval",
      "type": "local",
      "title": "Try shared-space & Bayesian retrieval",
      "subtitle": "137 development / 105 historical test molecules",
      "score": "Δ selected = 0",
      "verdict": "Not promoted",
      "tone": "reject",
      "finding": "Historical C2 MRR: re-fitted baseline 0.541994; vector-only 0.187738; Bayesian-only 0.315916. Development selection chose zero new-channel weight, so the selected fusion equaled the baseline.",
      "meaning": "The infrastructure works, but this small MLP and approximate fragment model did not earn an added role in the scored control. These results do not reject all learned or Bayesian retrieval."
    },
    {
      "id": "graph",
      "type": "local",
      "title": "Learn graphs & collision-energy effects",
      "subtitle": "Earlier graph / energy branch · historical cohort",
      "score": "+0.000775",
      "verdict": "Gates failed",
      "tone": "reject",
      "finding": "Selected historical C2 fusion moved 0.541994 → 0.542769, with delta interval [−0.013320, +0.013368]. Graph-only MRR was 0.134778 versus matched MLP 0.215990. Energy-conditioned MRR was 0.352006 versus blind 0.347522; its delta interval also crossed zero.",
      "meaning": "Graph, energy, and selected-fusion gates all failed. C1 remained unchanged. The additional 128-molecule historical cohort was not an independent external benchmark. No new official submission was made."
    },
    {
      "id": "glacier",
      "type": "local",
      "title": "Use pretrained GLACIER fragments",
      "subtitle": "Latest lead · 63 exposure-screened test molecules",
      "score": "+0.151977",
      "verdict": "Promising pilot",
      "tone": "",
      "finding": "GLACIER with marginalized spectral likelihood improved MRR from 0.705714 to 0.857691 against the matched energy head; paired 95% delta interval [+0.052649, +0.247336]. Development selected GLACIER alone. First-place answers increased from 38/63 to 49/63.",
      "meaning": "The prespecified local comparison passed. This historically explored, narrow cohort contains only one COCONUT test molecule and is not representative C2 validation. Complete checkpoint training ancestry is unavailable. Independent validation is required; no production promotion, leaderboard gain, or 0.4+ score is established."
    }
  ]
}