Skip to content

fix(heldout): use coverage-neutral interval field names - #1093

Draft
seonghobae wants to merge 3 commits into
codex/declared-heldout-bootstrap-20260907from
codex/heldout-interval-key-names-20260907
Draft

fix(heldout): use coverage-neutral interval field names#1093
seonghobae wants to merge 3 commits into
codex/declared-heldout-bootstrap-20260907from
codex/heldout-interval-key-names-20260907

Conversation

@seonghobae

Copy link
Copy Markdown
Contributor

Stack

Draft successor of #1092 (5bc0cd52). Parents #1091, #1090, #1074, and psychometric KPI owner #1067 stay open. This PR does not close #1092.

Exact head: d40a4981a258b4d0194bda3cdb6c99ae3097da50.

Problem

After declared coverage, nested held-out JSON fields still named *_ci95 implied 95% even when the run declared a different interval.

Change

  • Rename delta_ci95, paired_delta_ci95, query_delta_ci95, accuracy_delta_ci95, and heldout_paired_delta_ci95 to *_interval.
  • Coverage stays in declared bootstrap_confidence_level.
  • IRT interval_95_coverage_rate diagnostics are unchanged.
  • ADR 0045 is Proposed. Production route/conduct defaults stay locked.

Verification

Local smallest tests:

  • 20 passed in 3.53s (source-key contract, adaptive-calibration keys, declaration/boundary tests, ADR uniqueness)

This is naming-contract evidence, not buyer-held-out accuracy, p95 latency, or protected merge.

Remaining no-heuristics work

Other harness sample sizes remain later slices. No production policy or release is authorized.

RED: nested JSON fields still embed 95 in the name even though coverage
is a run declaration. The next commit must rename those fields.
Rename nested *_ci95 JSON keys to *_interval. Declared coverage stays in
bootstrap_confidence_level. IRT interval_95_coverage_rate diagnostics are
unchanged. Production route/conduct defaults stay locked.
ADR 0045 is Proposed. Gap baseline, changelog, and doctoring keep coverage
in the declared field rather than the JSON name.
@coderabbitai

coderabbitai Bot commented Sep 7, 2026

Copy link
Copy Markdown

Important

Draft PR not reviewed

Draft PRs are not automatically reviewed by default.

  • Trigger a manual review

To automatically review draft PRs, update your CodeRabbit configuration:

reviews:
  auto_review:
    drafts: true

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

Comment @coderabbitai help to get the list of available commands.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant