Round r65b

r65b

Status · INVALIDReviewed status is transcribed from rounds/STATUS.md. The original narrative and its qualifications remain below.

Source-reported narrativeRound README and results text are source material. Numeric values that are not tied to a claim ledger record remain source-reported.

Round source README · rounds/r65b/README.md

Round date: HISTORICAL (before 2026-10-09) Publication badge: INVALID; KEPT FOR AUDIT Why not VALID: The historical inventory retains INVALID because of NOT_SCORED, NO_PREREG, UNMATCHED_CONDITIONS. See the round evidence and limitations below. Recomputation status: OBSERVED SOURCE ONLY

Recomputation update: The mined raw records now provide per-cell outcome, manifest token-counter, and wall-time recomputation for the observed cohort. This does not reconstruct missing planned cells, a full intention-to-treat denominator, or model execution.

Status

INVALID

  • NO_PREREG — The README records no pre-registration.
  • NOT_SCORED — Official cells mark smoke rows not-scored while the prior table counted them as passes.
  • UNMATCHED_CONDITIONS — The documented format-gate confound and historical control harness/version differences violate matched conditions.

Required reproduction metadata

  • Kogen commit: `ce7b9dc7642b39d2a14c053a3814a7a5e2ca628d`, the round source's pinned build.
  • Harness commit: not recorded in the cited round source as a Git commit; any adapter or runtime fingerprints are not Git commits.
  • Model and effort: a complete per-cell mapping is not recorded in the cited round source; values present in the source summary remain as stated above.
  • Task IDs (source-listed plan): elx-02-ingest-supervision, elx-04-queue-backpressure, elx-07-cli-stats, elx-12-retry-api-deprecation, elx-port-board-publish-unpublish-public-boundary, elx-port-erase-account, syn-01-live-ticket-filters, syn-06-migration-ticket-numbers, syn-15-validation-ticket-form, syn-20-email-invite-flow, syn-24-csv-import, syn-31-inbound-email-webhook.
  • Historical execution command: not recorded in the cited round source.
  • Raw records: the cited sources do not provide a complete public round-specific request and grade bundle; public delivery and outcome exports are linked above where applicable.

Terminology: public round glossary.

STATUS: INVALID

Data completeness: 59.1% (90 deliveries; PASS WITH DECLARED GAPS). Measurement contract; declared gaps.

Status reason: The official export marks smoke rows not-scored while the prior table counted them as passes, so the aggregate and table are removed. The known format-gate confound and historical control differences in harness/version remain documented.

Pre-registered: no

Lifecycle: Historical capture. Snapshot 2026-10-05; earlier reports may use earlier cutoffs.

Question: Do stop fixes make final Kogen recipes competitive?

Venue: Public host identifiers remain in the linked exports; planned task-to-host assignments and execution-environment details are not established.

Design: 88 scored cells (8 held-out tasks × 2 arms × 3 reps, plus 4 discriminators × 2 arms × 5 reps) and two unscored smoke deliveries. The official export marks the smoke rows not-scored. The implementation behind the retained label is unavailable in the public record.

Decision rule: none predeclared beyond fixed design and Round 58 ITT.

Arms: Exact recipe definitions are not recoverable from the public record; exported arm labels remain in the linked results sources.

Planned tasks in the surviving round note:

Verdict: Outcome interpretation is withheld because the public outcome sources do not currently support a single reconciled classification. The previous headline and reconciliation table are removed; no comparison claim is made.

Public outcome reconciliation status

The prior reconciliation table is removed because its run-record classifications or denominator do not match the official outcome export. Consult cells.jsonl for official outcome states and r65b.jsonl for captured deliveries; this page does not combine them without a cell-level crosswalk.

Recomputed from raw records

`recomputed.json`, mined cell records, and the token audit back observed official outcomes, per-arm pass counts/rates, manifest token counters, and wall medians. The source contains 90 rows (fail 46, pass 44); overall pass rate is 48.9% (44/90) among pass/fail grades. Per-arm detail and coverage counts are in the JSON.

Published-export outcome cross-check: 90/90 shared cell IDs match; 0/90 public-export rows have no mined source record. Exact IDs and results for unpaired rows are in recomputed.json. No per-cell outcome mismatches were found in the shared IDs. Token reconciliation under METHOD.md: 90/90 published metadata totals equal the manifest input + cached input + output sum; 0 differ (no mismatches). Reasoning remains a separate reported component and is not added to output. The recompute keeps both numbers visible. 0 unexplained single-response numeric errors were identified; multi-response rows retain the published full-cell total and the manifest counter separately because per-response usage records are excluded.

Captured run records

Standard-schema records assigned to this round in the indexed snapshot.

90 records · 538,510 bytesDownload r65b.jsonl · View immutable source ↗

Reproduction inventory

Published analysis replay and full experiment replay are reported separately.

Replay type
unavailable
Script
No per-round analysis script is published.
Inputs
No identified public input bundle.
Full execution replay
Unavailable
Scope
No public per-round analysis script with a complete, identified input bundle is published. See the round README, MEASURED.md, and MISSING.md for recorded scope and gaps; historical execution inputs/toolchains/grading are not bundled.

View the complete per-round reproduction inventory · Repository reproduction guide