Finland

A bounded replay experiment with a dated evidence boundary.

Why Finland?

Finland combines open source material with difficult amendment histories. The enacted publications in the Statute Book of Finland are the source evidence for replay; Finlex also publishes consolidated texts with later amendments incorporated. That separation makes the consolidated text a comparison surface rather than the replay source.

When replay and the comparison surface differ, the mismatch has no automatic verdict. Possible explanations include a LawVM defect, missing or malformed source material, an editorial convention, a date-selection error, or a candidate issue for publisher review.

Claim boundary: this page reports dated research results. Current-corpus certification, source-chain closure, and official-error status require their own evidence and institutional disposition.

Explore a bounded timeline

The LawVM Explorer renders a Finnish statute for the dates and source history included in its manifest. It exposes point-in-time structure, provision provenance, and citation hashes. It is an external companion-domain research viewer, not an official consolidation or part of lawvm.org’s evidence boundary.

Open external Finnish research viewer Open external UK example

What the 0.23% figure means

Measured

0.23%
Project-reported mean normalized text edit distance from the declared Finlex comparison surface in an August 2026 development run.

Interpretation

Text distance
Use 0.23% as mean normalized distance to the declared Finlex comparison surface. Correctness and completeness require source review and residual classification.

Text distance mixes punctuation, notes, editorial conventions, source defects, replay defects, and potentially meaningful differences. Each residual needs its own evidence and disposition.

Finland review is public only in aggregate

A 2026 review tranche recorded 22 candidate findings that passed the project’s internal review threshold and were reported for external review. Each remains a candidate pending external adjudication. Public material stays aggregate-only until an individual record completes the disclosure and disposition process.

The public evidence ledger therefore records the Finland work at aggregate level. A case should become an individual public record only when its sources, disposition, publication boundary, and reproduction identity are suitable for independent review.

Next evidence gates

  • Freeze and publish a citable source-and-result bundle for a bounded tranche.
  • Rebuild the benchmark from documented inputs and current commands.
  • Classify every residual without presuming the comparison surface is authoritative.
  • Publish individual cases only after an appropriate review and disclosure decision.
  • Turn confirmed LawVM defects into named regression tests.
Assess a Finland pilot Inspect benchmark methodology Read the assurance limits
← LawVM Evidence Pilots