/research-status: 'Standing predictions' section — posture and pointers only (dated DOIs, self-grading, losses published); no research claims restated on-site

[emooreatx]

What

A short "Standing predictions" section on /research-status (linked from /proof's Theory class) that states the research posture and points off-site — deliberately without restating any research claim on the site itself.

Draft copy:

We make our research claims where they can be checked. Every major claim of this project is a dated, publicly archived prediction with a stated way it can fail — and, where possible, a script that grades it automatically, with no human vote. If a prediction fails, we publish the failure under the same names. The archive of current predictions (frozen 2026-07): [Zenodo DOI 10.5281/zenodo.21316928]. The framework paper: [Zenodo concept DOI 10.5281/zenodo.20300773]. The grading scripts and the full research record: [github.com/CIRISAI/coherence-ratchet].

We do not summarize the claims here, on purpose: a claim you cannot check where you read it is advertising. The links are the claims.

Why the pointer-only design (site policy — see the editorial-policy issue)

The site speaks to readers who cannot verify derivations; research content stays where the verification apparatus lives (Zenodo/GitHub). The site may state the posture — public, dated, self-grading, losses published — because the posture is checkable culture: a reader can follow the DOI and confirm the dates themselves.

Notes

• Placement: /research-status, with /proof's Theory card pointing at it (existing anchor structure works).
• No physics vocabulary on the page. No claim names. Dates + DOIs + the loss-publication sentence only.
• 29 locales.