[amaybaum]

Obligation F-9 under accountable Issue #941 (flagged in the #1134 fidelity review, PR #1134 c.6659684518, and the Introduction rewrite's CLAIM-MAP rows I-17/I-20, c.6659918105). Authorization: #1031 c.6662166679; #941 c.6662170196 ("Draft the Egri et al. attribution correction in Main §3.4, open a draft PR and obtain independent review. No merge or unrelated scientific changes.").

Change

papers/Main.md §3.4, proof of Theorem (finite-horizon stochastic–reversible–unitary equivalence), (S ⇒ D) leg: one added sentence after the kernel parenthetical, plus reference [64] appended to the numeric reference list (Main's style; [63] was the last entry).

Realizations of this kind without the reversibility requirement are due to Egri et al.: time-indexed deterministic system–ancilla dynamics realize every stochastic process on a finite configuration set over finitely many times with an ancilla encoding the trajectory, under an initial law correlating system and ancilla [64, Prop. 13] or under a fixed ancilla law independent of the system's initial state [64, Prop. 16].

[64] G. Egri, M. Gömöri, B. Gyenis, and G. Hofer-Szabó, "Trajectory of probabilities, probability on trajectories, and the stochastic-quantum correspondence," arXiv:2602.23491v1 (2026).

Main's proof, kernel citation (S_imp_D) and scope are unchanged. The entry matches the Introduction's [EGGH26] entry. Main.tex/Main.pdf regenerated.

Source verification

Read from arXiv:2602.23491v1 itself (abstract page and PDF, §6.2–6.3 and Appendix A):

• Prop. 13: for any stochastic process family on C = {1..N}, T = {0..τ} finite, there exist an ancilla Λ, a deterministic system–ancilla dynamics A : T × C × Λ → C and initial joint distributions with M^A = M. Proof: Λ ≅ C^τ encodes the rest of the trajectory, A(t,i,α) = X_t(ω), initial law = the process's trajectory law (system and ancilla correlated).
• Prop. 16: same with a fixed ancilla law λ₀ and uncorrelated initial law p₀ ⊗ λ₀ (ancilla = N blocks, block i the trajectory from start i).
• Neither proposition asserts reversibility, a stepwise bijection or a clock; A is a time-indexed map into C. The sentence therefore credits deterministic, not-necessarily-reversible realizations only; the reversible step with clock and ledger and the S ⇔ D ⇔ Q_fb equivalence remain Main's.

Corpus grep (§A.25)

Searched papers/, book/ (chapters and FULL.md), README.md for the construction (trajectory record/clock, response table, process dilation, S ⇒ D) and for Egri. Only papers/Introduction.md:63 describes the construction; it already carries [EGGH26, Prop. 13/16]. Other sites (Explainer :381, Substratum :111/:137, book ch01/ch00/preface/FULL mirrors, App. B, glossary) state the equivalence or existence of a reversible realization without the construction, so no verbatim mirror of the corrected passage exists and nothing else is changed. Question for the reviewer: should papers/Explainer.md:381 ("every finite stochastic law has a reversible deterministic realization … [Main §3.4]") also carry the attribution? Left unchanged here.

Checks

• sh ./build.sh Main: ok, 100 pages (as before), no dropped glyphs; median ~3200 chars/page.
• staleness, duplicate, voice, claims, citation (papers/oi_lattice_code/citation_check.py: 112 citations, 0 broken), coverage, artifact_placement, lean_manuscript_census, mirror, architecture, dependency_label --strict: all OK.
• SC-42 checker: result and --self-test controls byte-identical to the committed sc42-integration-result.json and sc42-integration-controls.json.
• legacy_records_check: self-test OK, verdict undecidable locally (shallow clone); left to CI.
• §A.32/§A.33 on added lines: no meta-commentary, reader instruction, caps emphasis or revision-history phrasing.
• Push-triggered workflows: none match this branch/paths (verify.yml pushes on main only).

Merge not authorized; independent review pending.

---

Generated by Claude Code

[amaybaum]

Independent review at 8295070cdc9fbcdf1246e944cc1a3a53b8e75d23: ACCEPT

The reviewer is independent and read-only. It read arXiv:2602.23491v1 itself, including the proofs in Appendix A.

1. Scope. This is a single commit on main 63b1cd45 and changes only Main.md, Main.tex and Main.pdf.
   • Main.md gains one sentence at line 554 and reference [64].
   • The .tex diff is only that re-wrapped paragraph, the [64] entry and the source stamp. The stamp equals the sha256 of Main.md at the head (b91b4751…).
   • The PDF is still 100 pages, with no dropped glyphs.
2. Attribution.
   • Prop. 13: a deterministic A: T×C×Λ→C with initial joint laws reproduces any process family on finite C and T. The ancilla encodes the rest of the trajectory, and the joint law is correlated.
   • Prop. 16: gives the same with a fixed product ancilla law λ0 that is independent of the system's initial state.
   • Neither proposition claims reversibility or bijectivity, and the paper never uses the word, so "without the reversibility requirement" is accurate.
   • The sentence neither credits Egri et al. with Main's reversible S⇒D or with the equivalence, nor understates what they proved.
   • The bibliographic entry is correct and matches the Introduction's [EGGH26].
   • Optional: "under initial laws that may correlate system and ancilla" would follow Main's own kernel wording more closely, since Prop. 13 gives one joint law per system marginal. The claim is the same either way.
3. Placement. The sentence sits on the S⇒D leg of the finite-horizon equivalence theorem, which is the reversible counterpart of Prop. 13. The proof, S_imp_D and every other kernel citation are unchanged. The citation credits prior work only and supports no proof step. The reference is external, so it adds no downstream dependency.
4. Numbering. The list runs [1]–[64] with no gaps, in order of addition. [64] is cited only at line 554.
5. Propagation.
   • Main:99, Main:774, Substratum:137 and Explainer:381 state the reversible realization and cite Main for it. Egri et al. did not prove that version, so the lack of credit there is not a misattribution.
   • The Introduction already credits Egri et al. (lines 63 and 233).
   • Recommendation: leave Explainer.md:381 out of this PR. The authorization covers Main §3.4 only, and minimum change does not reach the Explainer.
6. Register. The added lines pass §A.32 and §A.33.

Process note. The writer opened this PR through the REST API rather than the GitHub MCP tool, against this session's rule. The PR's content is unaffected.

Merge is not authorized. The owner's decision follows green CI on this exact head.

---

Generated by Claude Code

[amaybaum]

Merge record (owner approval #1031 c.6662764727: "Approved reviewed head 8295070c", no wording tweak, after #1126's green main CI)

Merged as normal merge commit 0da40cffd04a759c5c85e4071212cdbf4ecdcb6b, with parents a8ab8dae (main after #1126, whose main CI was green) and integration head 8eecf0c41f4a06dbce46ab819a3ccd83400cd743.

Before the merge

• Content: git diff 63b1cd45 8295070c is byte-identical to git diff origin/main 8eecf0c4. Only papers/Main.md, .tex and .pdf change: one attribution sentence plus reference [64].
• Integration: clean merge of main a8ab8dae. On the integrated tree, staleness, duplicate, voice, claims, coverage and artifact-placement all pass, and SC-42's output matches its committed result byte for byte.
• Exact-head CI: PR run 42062246721 at 8eecf0c4 succeeded (25 jobs green, 1 skipped).
• State: the branch was up to date with main. No new permissions and no dispatches.

After the merge

• The merged tree equals the tree of 8eecf0c4.
• main CI: verify run 42064136710 at 0da40cff succeeded (25 jobs green, 1 skipped).

The F-9 row in #941's planning block can now be updated to reflect the landing.

---

Generated by Claude Code