Reproducibility

Research cut-off date: 2026-07-22. Status: Phase 2, not yet written. This file states the scope of the work, the arguments it owns, and the research required to complete it. It contains no findings, because none have been sourced. See the changelog for phase status.

Scope

Specifies what must be recorded for any evaluation in this repository to be repeatable: model identifier and version, sampling parameters and seed, harness and version, prompt templates, date and region, and the full output artefacts. Also states the limits: endpoints change without notice, and some non-determinism cannot be eliminated by seeding.

Research checklist

• Specify the complete run record and where it is stored.
• State which sources of non-determinism can be controlled and which cannot, with evidence for the latter.
• Specify the re-verification interval after which a recorded result is treated as historical rather than current.
• Specify the archival policy for prompts and outputs, including any redaction required before contribution.

Completion criteria

This file is complete when every checklist item above is closed and the quality-control checklist in the research methodology passes.