Answers are grounded in this paper and cite their sources.
UNREVIEWED RESEARCH ARCHIVE / NOT PEER REVIEWED. This supplement connects two distinct observation contracts to the previously published Couret source archive (10.5281/zenodo.23980513) and unified manuscript (10.5281/zenodo.23981308). It contains 221 files, including two readable French HTML notes, formal sources, exact certificates, historical compilation evidence and fresh replay results. Four-point reconstruction: version 7 studies the complete fibre of a four-point subset of a finite cyclic group under its labelled ordered-convolution square, including diagonals. The fibre is affine over F2 with dimension d at most two. Fixed membership probes and a mathematical decoder attain worst-case additional query cost d; every correct deterministic adaptive membership tree has depth at least d. The cost excludes acquisition of the square and computation of the fibre. The 13 new modules and all 21 inherited modules are included, with exact source locks. The historical local Lean 4.29.0 receipt covers 298 audited roots, 34 modules and 36 successful commands. No fresh Lean compilation is claimed for this publication; no executable implementation or running-time guarantee is certified. Predictive observations: on two explicitly defined directed Cayley graphs on C12, the half-turn quotient is autonomous, whereas observing mass at vertex zero and its complement requires ten additional linear observables to close prediction for all initial distributions, with generator known. These are dimensions of linear observables, not bit counts. Exact rational ranks and closure identities are separate from double-precision trajectory checks. General observability theory is classical. Fresh publication checks: consistency of the finite receipts, three boundary witnesses, eight verifier tests, all 37 locked source files, the 298-root historical axiom log, and a full PA03 replay passed. The C4..C48 exhaustive campaigns were not rerun. Novelty and independent human review are not established. Prepared with substantial AI assistance. The universal reconstruction proof is a distinct later dossier relative to the finite-case discussion in the unified manuscript; it does not compile or validate that manuscript's DecisionInterface.lean. General counting formulas remain open in Lean. No RH result, physical prime law or FCI advantage is asserted. No joint Koyama manuscript, private correspondence, full historical facsimiles or confidential FCI files are included. Rights: no new blanket reuse licence is granted; existing explicit notices retain their scope. See README.txt and provenance files.
No takes yet. Share an insight, caveat, or question.
Alexandre Couret (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: