Skip to content

fix(docs): drop Recital 133 misattribution from conformal explainer#125

Merged
vaaraio merged 1 commit into
mainfrom
fix/docs-recital-133-misattribution
May 22, 2026
Merged

fix(docs): drop Recital 133 misattribution from conformal explainer#125
vaaraio merged 1 commit into
mainfrom
fix/docs-recital-133-misattribution

Conversation

@vaaraio
Copy link
Copy Markdown
Owner

@vaaraio vaaraio commented May 22, 2026

Recital 133 of the EU AI Act is about synthetic content watermarking under Article 50, not detection of non-conformities or accuracy verification. The conformal-prediction.md doc and v0.28.0 CHANGELOG entry both cited it incorrectly.

Verified the actual Recital 133 text via artificialintelligenceact.eu before this fix. The recital states: "A variety of AI systems can generate large quantities of synthetic content that becomes increasingly hard for humans to distinguish from human-generated and authentic content... it is appropriate to require providers of those systems to embed technical solutions that enable marking in a machine readable format and detection that the output has been generated or manipulated by an AI system."

That maps to Article 50 (transparency for synthetic content) and provider watermarking obligations, not to Article 15 accuracy/robustness or runtime detection of operational non-conformities.

Fix

  • docs/conformal-prediction.md: drop the Recital 133 paragraph. Article 15(1) anchor stands on its own. Replaced the recital framing with a tighter explanation of why a widening interval is itself the real-time signal.
  • CHANGELOG.md v0.28.0 entry: drop the "and Recital 133 (detection of non-conformities)" clause from the conformal-prediction.md description.

Why this is going in before tagging v0.28.0

v0.28.0 is merged to main but not yet tagged and not yet on PyPI. Landing this fix on main first means the tag points at the corrected content and PyPI never publishes the misattribution. Cleaner than v0.28.1 a few minutes later.

Tests

771 passed, 12 skipped. No code change.

Summary by CodeRabbit

v0.28.0 Release Notes

  • Documentation

    • Added plain-language explainer for conformal prediction and risk score intervals
    • Added public reference documenting evidence-sufficiency verdict logic, decision thresholds, and how auditors interpret compliance reports
    • Updated compliance documentation with clearer threshold definitions
  • Chores

    • Version bumped to 0.28.0

Review Change Stack

@coderabbitai
Copy link
Copy Markdown

coderabbitai Bot commented May 22, 2026

Warning

Rate limit exceeded

@vaaraio has exceeded the limit for the number of commits that can be reviewed per hour. Please wait 33 minutes and 51 seconds before requesting another review.

You’ve run out of usage credits. Purchase more in the billing tab.

⌛ How to resolve this issue?

After the wait time has elapsed, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

We recommend that you space out your commits to avoid hitting the rate limit.

🚦 How do rate limits work?

CodeRabbit enforces hourly rate limits for each developer per organization.

Our paid plans have higher rate limits than the trial, open-source and free plans. In all cases, we re-allow further reviews after a brief timeout.

Please see our FAQ for further information.

ℹ️ Review info
⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: a2b04382-1a05-4980-af0e-60697fd702c5

📥 Commits

Reviewing files that changed from the base of the PR and between 4d61f6a and fe1b7f3.

📒 Files selected for processing (2)
  • CHANGELOG.md
  • docs/conformal-prediction.md
📝 Walkthrough

Walkthrough

This PR releases v0.28.0 with two new documentation artifacts: VERDICTS.md (a public reference for evidence-sufficiency thresholds and compliance-engine decision logic) and docs/conformal-prediction.md (a non-technical explainer for conformal prediction intervals). Existing documentation is updated with cross-links, and project versions are synchronized to 0.28.0.

Changes

v0.28.0 Documentation Release

Layer / File(s) Summary
Evidence-Sufficiency Threshold Reference
VERDICTS.md
VERDICTS.md documents the complete evidence-sufficiency decision logic, including status/strength taxonomy (evidence_sufficient, evidence_partial, evidence_insufficient, not_applicable), the engine's decision tree, per-article threshold tables for EU AI Act and DORA, and auditor-visible JSON examples showing chain-integrity pinning.
Conformal Prediction Interval Explainer
docs/conformal-prediction.md
New document explains conformal prediction as a distribution-free uncertainty quantification method, including interval properties, connections to EU AI Act Article 15(1), and clarification that intervals supplement (not replace) ground-truth outcome tracking.
Documentation Cross-Linking and Index Updates
CHANGELOG.md, COMPLIANCE.md, README.md
CHANGELOG.md introduces v0.28.0 release notes describing the new docs and threshold defaults; COMPLIANCE.md adds a pointer to VERDICTS.md; README.md adds docs/conformal-prediction.md to the "Where things live" index.
Project Version Synchronization
pyproject.toml, clients/ts/package.json, src/vaara/__init__.py
Version constants bumped from 0.27.0 to 0.28.0 across Python package, TypeScript client, and main module.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

Possibly related PRs

  • vaaraio/vaara#16: Both PRs update COMPLIANCE.md documentation to describe how ComplianceEngine EU AI Act/DORA article evidence rules map to runtime outputs; this PR adds a pointer to VERDICTS.md for threshold details.
  • vaaraio/vaara#95: Both PRs modify the "Where things live" documentation index in README.md, reorganizing and adding entries to the docs table.

Poem

🐰 Release day hops with clarity bright,
Evidence thresholds now plainly in sight,
Conformal intervals explained—no more fright!
Cross-links and versions aligned just right,
Version point-two-eight, a documented delight! ✨

🚥 Pre-merge checks | ✅ 5
✅ Passed checks (5 passed)
Check name Status Explanation
Description Check ✅ Passed Check skipped - CodeRabbit’s high-level summary is enabled.
Title check ✅ Passed The PR title clearly and specifically describes the main fix: removing an incorrect attribution of Recital 133 from the conformal prediction documentation, which matches the stated objectives and the actual changes made.
Docstring Coverage ✅ Passed No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check ✅ Passed Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check ✅ Passed Check skipped because no linked issues were found for this pull request.

✏️ Tip: You can configure your own custom pre-merge checks in the settings.

✨ Finishing Touches
🧪 Generate unit tests (beta)
  • Create PR with unit tests
  • Commit unit tests in branch fix/docs-recital-133-misattribution

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 and usage tips.

Recital 133 of the EU AI Act is about synthetic content watermarking under Article 50, not detection of non-conformities or accuracy verification. The conformal-prediction.md and v0.28.0 CHANGELOG entry both cited it incorrectly. Dropping the recital reference; the Article 15(1) anchor stands on its own and is verified against the AI Act text.
@vaaraio vaaraio force-pushed the fix/docs-recital-133-misattribution branch from 4d61f6a to fe1b7f3 Compare May 22, 2026 16:51
@vaaraio vaaraio merged commit 7ba9b21 into main May 22, 2026
10 checks passed
@vaaraio vaaraio deleted the fix/docs-recital-133-misattribution branch May 22, 2026 16:55
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