OP-48(c) iteration-7 attack: the full-pair reading is forced — horn C loses its definitional residue; plus the OP-50 watch

Status: ATTEMPT executed — didItMove = residue-resolved on the OP-48(c) definitional question (Part 1) and watch-updated, three-cornered on OP-50 (Part 2). Medium-high confidence. The headline verdict of ../CONCLUSION.md is unaffected and tightened. Last updated: 2026-06-10 Iteration: 7 (track A4)

This note executes the iteration-7 named decidable step left by 2026-06-10-iter6-op44-op48c-levers.md §2.5–2.6: produce a principled argument for or against the full-pair formalization of "recovers QM" — the definitional residue keeping horn C of ../OPEN_PROBLEMS.md OP-48(c) from a theorem — by reading the actual axiom statements of the three GPT reconstruction theorems. Plus a quick verified sweep of the OP-50 closed-universe dispute (post-2026-01).

Scope honesty, up front. (1) All three reconstruction papers were read at the primary-PDF level this session for the axiom and framework statements quoted below; page numbers refer to the arXiv PDFs (Hardy v4; CDP v3; MM v4). Journal references for CDP (PRA 84, 012311) and MM (NJP 13, 063001) are standard but were not independently verified this session [unverified]. (2) The operator-algebraic inputs (product-state/split equivalence; Summers–Werner) are carried from iterations 4–6 with their standing hedges; nothing here re-derives them. (3) The Step-2 separation lemma is the author's own elementary derivation; the referee should check the density argument.

Part 1 — The two readings, formalized and tested against the actual axioms

1.1 The residue as iteration 6 left it

Iter-6 hardened horn C to "theorem-shaped under the full-pair reading, vacuous under the subsystem reading," leaving as residue the choice between readings:

(FP) Full-pair: the GPT systems are $A = (\\mathcal M)_*$ , $B = (\\mathcal M')_*$ (normal states, Haag-dual pair, $\\mathcal M$ Type III $_1$ ), composite $AB = (B(\\mathcal H))_*$ ; "recovers QM" = the composition axioms hold on this triple.
(SS) Subsystem: "recovers QM" = some subfactor pair $F_A \\subset \\mathcal M$ , $F_B \\subset \\mathcal M'$ satisfies the axioms on $(F_A, F_B, F_A \\vee F_B)$ .

1.2 What the reconstruction theorems actually say `[ESTABLISHED — primary texts read this session]`

Hardy (quant-ph/0101012 v4). Axiom 4, verbatim (p. 2): "Composite systems. A composite system consisting of subsystems A and B satisfies $N = N_A N_B$ and $K = K_A K_B$ ." The operational unpacking (§8.8, Fig. 2, p. 24) quantifies over the composite produced by the given preparation device: the $K_A K_B$ fiducial measurements are product measurements ("we simply perform the $i$ th fiducial measurement on A and the $j$ th fiducial measurement on B"), and the projectors $\\hat P_i^A \\otimes \\hat P_j^B$ correspond "to preparing the $i$ th fiducial state at side A and the $j$ th fiducial state at side B" — independent product preparations of the designated pair are presupposed.

Chiribella–D'Ariano–Perinotti (arXiv:1011.6451). Two layers, cleanly separated in the paper:

Framework (§II, p. 5): "From every pair of systems A and B one can form a composite system, denoted AB"; and Eq. (3): for experiments run in parallel "we assume that the joint probability distribution is given by the product." Product states $\\rho \\otimes \\sigma \\in \\mathrm{St}(AB)$ exist by framework definition, before any axiom. P. 6 restricts to finite-dimensional state spaces.
Axiom (p. 10), verbatim: "Axiom 4 (Local distinguishability). If two bipartite states are different, then they give different probabilities for at least one product experiment." Quantified over $\\rho, \\sigma \\in \\mathrm{St}_1(AB)$ — the given composite. The paper states the equivalence with local tomography, $D_{AB} = D_A D_B$ , and the product-basis expansion $\\rho = \\sum_{ij} \\rho_{ij}\\, \\alpha_i \\otimes \\beta_j$ ; the $D_A D_B$ count consumes the framework's product states.

Masanes–Müller (arXiv:1004.1483 v4). Framework §IID (pp. 3–4): composite defined via joint measurements + no-signaling; verbatim: "This implies that all product states $\\psi_{AB} = \\psi_A \\otimes \\psi_B$ are contained in $\\mathcal S_{AB}$ ." Requirement 2 (p. 5), verbatim: "(Local tomography). The state of a composite system AB is completely characterized by the statistics of measurements on the subsystems A, B," equivalent to $(d_{AB}+1) = (d_A+1)(d_B+1)$ .

Reading off the answer to the mission's question: local tomography quantifies over THE pair in all three formalizations. None contains an existential quantifier over embedded pairs. The SS reading is not a reading of any of these axioms: it can only be implemented by deleting the full pair from the theory's system roster — which abandons OP-48's question (whether composition axioms can run on the subregion algebras themselves) and collapses into the iter-6 §2.4 vacuity (any SS witness is split with $N = F_A$ Type I). FP is the only non-vacuous formalization. It is forced. [INFERENCE, high]

1.3 Decomposition: where exactly the III $_1$ pair fails (own derivation)

Split the composition package into (PS) product-preparation primitive (framework level), (LT-sep) separation by product experiments (CDP Axiom 4 verbatim), (LT-dim) the dimension count / product-basis expansion. On the III $_1$ Haag-dual pair under FP:

(LT-sep) HOLDS. $\\mathrm{span}\\{ab : a \\in \\mathcal M, b \\in \\mathcal M'\\}$ is a unital $*$ -subalgebra (commutation gives $(ab)(a'b') = (aa')(bb')$ ), $\\sigma$ -weakly dense in $(\\mathcal M \\cup \\mathcal M')'' = B(\\mathcal H)$ (factor; von Neumann density theorem). Two normal states agreeing on a $\\sigma$ -weakly dense subspace agree everywhere; decomposing algebra elements into effects, distinct normal states of $AB$ differ on some product experiment $\\varphi(ab)$ . So the verbatim CDP local-distinguishability axiom is satisfied by the Type III $_1$ pair. [INFERENCE — elementary, mine]

(PS) FAILS, uniformly. Carried from iter 4–6: a normal product state across $(\\mathcal M, \\mathcal M')$ forces $\\mathcal M$ Type I (T1 + Haag-dual corollary); quantitatively every normal state is norm-distance $\\ge 1 - 1/\\sqrt 2$ from every separable state (Summers–Werner + R6-F6 gap, with the standing primary-source hedge).

(LT-dim) FAILS trivially (infinite dimension; and its lower bound consumes (PS) anyway).

Consequence. The failure locus of horn C is not tomography — it is the product-preparation primitive (PS), which sits at framework level in all three reconstructions, upstream of their tomography axioms. The mission's conditional ("if CDP-style local tomography requires product states on the given bipartition, FP is forced") resolves sharper than posed: separation-style local tomography does not require product states and even survives on III $_1$ ; the framework's parallel composition does require them, on the given bipartition, in all three papers.

1.4 Horn C, final form

Statement. Let $T$ be an operational theory whose system roster includes $A = (\\mathcal M)_*$ , $B = (\\mathcal M')_*$ , $AB = (B(\\mathcal H))_*$ , $\\mathcal M$ Type III $_1$ with separable predual. Then $T$ violates (PS) — common to Hardy 2001 / CDP 2011 / Masanes–Müller 2011 at framework level — with uniform gap $1 - 1/\\sqrt2$ ; a fortiori $T$ models none of the three axiom sets. Conversely (PS) on the pair forces $\\mathcal M$ Type I.

Conditional on: (i) the commuting-pair product-state/split equivalence (carried [ESTABLISHED], Doplicher–Longo 1984 et al.); (ii) Summers–Werner 1988 (carried spot-check hedge); (iii) physical states = normal states. Not conditional on a choice of reading — that freedom is gone. This is theorem-conditional-on-framework, which subsumes and strengthens the target "theorem-conditional-on-CDP." [INFERENCE, high]

The two named exits (the relocated residue):

(X1) Singular states. Non-normal product states across the pair exist (product states on $\\mathcal M \\otimes_{\\max} \\mathcal M'$ push to $C^*(\\mathcal M \\cup \\mathcal M')$ by the universal property for commuting representations). Admitting them as preparations evades horn C but abandons density matrices and $\\sigma$ -additivity — i.e. relaxes the program's own state postulate. The encodes-not-generates installation pattern recurs at the state-postulate layer.
(X2) Non-OPT composition. The III $_1$ /universal-embezzlement composition endpoint (iter-6 §2.5, as corrected by R6-F7) — exits the operational-probabilistic framework entirely. Unchanged; operator algebra cannot choose.

Part 2 — OP-50 watch (post-2026-01, all verified at abstract level this session)

Harlow–Usatyuk–Zhao (arXiv:2501.02359) now published, JHEP 02 (2026) 108 — journal milestone for the observer-relative $e^{S_{Ob}}$ camp. [ESTABLISHED, bibliographic]
Nomura–Ugajin, Physical Predictions in Closed Quantum Gravity (arXiv:2602.13387; v1 Feb 13, v2 Mar 1, 2026): accepts per-sector one-dimensionality and builds a framework in which observers' limited access suppresses fluctuations, yielding statistically stable semiclassical predictions (Hartle–Hawking input). [ESTABLISHED, abstract-verified]
Higginbotham, Helping observers in closed universes reach their full potential (arXiv:2512.17993; revised Jan 8, 2026; JHEP 03 (2026) 183): refined SWAP-test operators on the Antonini–Rath puzzle; the AdS observer cannot rule out a baby universe, softening the Engelhardt–Gesteau–Harlow observer-complementarity tension. [correction: sole author Higginbotham, building on EGH — an initial search snippet misattributed it] [ESTABLISHED, abstract-verified]
Miyata–Horikoshi–Tajima–Ugajin, Emergent Closed Universes in Symmetric Orbifold CFTs (arXiv:2606.04575, Jun 3, 2026): closed-universe sectors whose physical Hilbert space grows only polynomially in $N$ after gauge constraints; pure states indistinguishable from mixed; sector naturally hyperfinite Type II $_1$ — a third position, neither strictly 1-dim nor $e^{S_{Ob}}$ , cross-linking to OP-44's II $_1$ structures. [ESTABLISHED, abstract-verified]
Negative sweep [dated process note, 2026-06-10]: no new Engelhardt–Gesteau paper post-Jan-2026; no direct ARSSV or Liu reply in the window (Liu-framework paper arXiv:2509.14327 and EGH arXiv:2507.06046 are both pre-window, Sep/Jul 2025). Search-bounded.

OP-50 verdict: stays [CONTESTED]; the dispute is now three-cornered. No bearing on the standing wiki verdict.

Honest overall verdict

Part 1 is a positive resolution of a named decidable step: the definitional residue of horn C is eliminated, not by fiat but by reading the axioms — the subsystem reading is inexpressible in any of the three reconstruction systems, and the load-bearing failure is a framework primitive shared by all three. The surprise sub-finding (separation-tomography holds on III $_1$ ) is a genuine sharpening: horn C was never about tomography; it is about preparability. The residue relocates to (X1) the normal-state postulate and (X2) the composition-primitive choice — both installation choices, consistent with HYP-ENCODING-SCREEN. Part 2 is maintenance: three verified developments, one attribution correction, one new structural datum (polynomial-dim II $_1$ sector) that feeds OP-44 rather than OP-50's dichotomy.

Open subquestions (updated)

[OPEN — carried] Primary-source spot-check of Summers–Werner 1988 hypotheses (iter-6 #5) — now the only mathematical hedge left under horn C.
[OPEN — new] Is there a reconstruction-grade axiomatization of bipartite composition that replaces product preparations with a preparability-free primitive (e.g. embezzlement-monotone constraints) and still singles out a unique theory? A positive answer would make the (X2) exit a genuine rival axiomatization rather than a gesture.
[OPEN — new, OP-50/OP-44 crossover] Does the 2606.04575 polynomial-dimension II $_1$ sector survive coupling to an observer in the Harlow–Usatyuk–Zhao sense, or does it collapse to one of the two standing camps?

References

Primary-PDF read this session (axiom/framework statements quoted with page numbers) unless marked otherwise.

L. Hardy, Quantum Theory From Five Reasonable Axioms, arXiv:quant-ph/0101012 v4. [PDF read: Axiom 4 p. 2; §8.8 + Fig. 2 pp. 24–25]
G. Chiribella, G. M. D'Ariano, P. Perinotti, Informational derivation of Quantum Theory, arXiv:1011.6451. [PDF read: framework §II Eq. (3) p. 5, finite-dim p. 6; Axiom 4 + local-tomography equivalence p. 10] Phys. Rev. A 84, 012311 (2011) [journal ref unverified this session].
L. Masanes, M. P. Müller, A derivation of quantum theory from physical requirements, arXiv:1004.1483 v4. [PDF read: §IID pp. 3–4; Requirement 2 p. 5] New J. Phys. 13, 063001 (2011) [journal ref unverified this session].
M. Plávala, General probabilistic theories: An introduction, arXiv:2103.07469. [abs fetched; the min/max-tensor composite statement NOT verified verbatim — corroborative only]
D. Harlow, M. Usatyuk, Y. Zhao, Quantum mechanics and observers for gravity in a closed universe, arXiv:2501.02359; JHEP 02 (2026) 108. [abs + Springer page via search]
Y. Nomura, T. Ugajin, Physical Predictions in Closed Quantum Gravity, arXiv:2602.13387 (v2 Mar 1, 2026). [abs fetched]
K. Higginbotham, Helping observers in closed universes reach their full potential, arXiv:2512.17993 (rev. Jan 8, 2026); JHEP 03 (2026) 183. [abs fetched; Springer via search]
A. Miyata, K. Horikoshi, H. Tajima, T. Ugajin, Emergent Closed Universes in Symmetric Orbifold CFTs, arXiv:2606.04575 (Jun 3, 2026). [abs fetched]
Carried, not re-verified this session: Doplicher–Longo, Invent. Math. 75 (1984); S. J. Summers, R. F. Werner, Ann. Inst. H. Poincaré 49 (1988) 215–243 [iter-6 hedge stands]; van Luijk–Stottmeister–Wilming, arXiv:2409.07646, Quantum 9, 1818 (2025); van Luijk, arXiv:2510.07563. Standard operator-algebra facts used without citation: von Neumann density theorem; max-tensor universal property [textbook, unverified].

Binding note. Where a correction in the referee verdict below conflicts with the body above, the referee correction governs; the body is the pre-referee submission, retained for the audit trail per house convention.

Referee verdict

Track A4-op48c-formalization-and-op50, iteration 7. All seven findings kept; one MAJOR correction (R7-A4-F5), one scope correction (R7-A4-F6), the rest minor or clean. Citation audit: every load-bearing source independently re-fetched this session — the three reconstruction papers were read page-by-page from primary PDFs and every verbatim quote is exact (Hardy Axiom 4 p.2 and §8.8 p.24 incl. the Appendix-4 projector correspondence; CDP composite-system sentence and Eq.(3) p.5, finite-dim restriction p.6, Axiom 4 with the D_AB = D_A D_B equivalence p.10; MM §IID product-state containment p.4, Requirement 2 and Eq.(8) p.5). No fabrications, no misattributions. The four OP-50 items all check out, including both journal refs absent from the arXiv abs pages (HUZ JHEP 02 (2026) 108; Higginbotham JHEP 03 (2026) 183 at Springer) and the submission's own Higginbotham attribution correction.

R7-A4-F1 — keep (minor). Quotes exact; no existential quantifier over embedded subfactors exists in any of the three axiom systems — confirmed by direct reading. Tag repair: the textual layer is ESTABLISHED, the forcing consequence is INFERENCE, high and must carry its own tag. Hardy admits countably infinite N; only CDP and MM state the finite-dim restriction verbatim.

R7-A4-F2 — keep (none). Product states are framework-level in all three, upstream of the tomography axioms, and the dimension-count proofs consume them — verified in the primary texts. The load-bearing-axiom identification (PS, not local tomography) is the track's correct structural insight.

R7-A4-F3 — keep (minor). The separation proof was re-derived independently and is correct (unital *-subalgebra of products; S' = M ∩ M' = ℂ1 for a factor; von Neumann density; normality; four-effects decomposition). The separation half of local tomography holds on the III₁ pair — the failure locus of horn C is exactly the product-preparation primitive. Gloss required: strictly, the III₁ triple is outside CDP's finite-dim framework; what holds is the dimension-free separation content of Axiom 4.

R7-A4-F4 — keep (minor). Theorem-conditional-on-framework upgrade legitimate (weaker shared premise ⇒ stronger conditional); conditionals (i)–(iii) honestly listed, the Summers–Werner primary spot-check hedge still open and still carried. Flag: 'models none of the three axiom sets' is overdetermined by finite-dimensionality alone; the non-trivial content is the (PS)-violation with the 1−1/√2 gap against LT-sep holding.

R7-A4-F5 — keep (severity: MAJOR, corrected). The X1 existence route is invalid as stated: the universal property gives a surjection M ⊗_max M' → C*(M ∪ M'), and states do not push forward through a surjection — φ_A ⊗ φ_B need not vanish on the kernel. Corrected route (binding): factor ⇒ Schlieder property ⇒ Roos embedding of M ⊙ M' in B(H) ⇒ the induced C*-norm dominates the min norm ⇒ φ_A ⊗ φ_B extends to a state on C*(M ∪ M') ⇒ positive Hahn–Banach extension to B(H). Conclusion (singular product states exist; admitting them relaxes the normal-state postulate; X2 exits the OPT framework; residue relocated, not eliminated) survives. The factor hypothesis is essential and may not be dropped.

R7-A4-F6 — keep (minor, corrected). All four watch items verified at abstract level and both journal refs confirmed. Scope correction on arXiv:2606.04575: the hyperfinite II₁ algebra and pure≈mixed indistinguishability are pre-gauge-constraint features of the exponentially large dominant sector; the polynomial-in-N dimension is the post-constraint physical Hilbert space. The OP-44 cross-link must cite the pre-constraint structure. Bonus omitted detail that strengthens the wiki's observer thread: Hartle–Hawking alone fails for the dominant sector and agreement is restored by coupling an external observer system.

R7-A4-F7 — keep (none). Negative sweep independently re-run and not contradicted; properly dated and search-bounded per EPISTEMICS.md. Optional enrichment: note pre-window arXiv:2509.14338 ('A no-go theorem for large N closed universes') in the OP-50 dossier.

No-go and closure audit (track level). Bell/CHSH/Tsirelson again used in the wiki's favor and correctly (carried R6-F6 gap; arithmetic re-checked). Reeh–Schlieder, Haag, PBR, Kochen–Specker, Coleman–Mandula, Weinberg–Witten, spin–statistics: not implicated. Wiki closures: dS-cardinality untouched; HYP-CKV-VACUITY-R3 and its 7-item smuggle list (incl. indefinite-pairing axiom and signature-valued template) not engaged by this track — checked, no quiet contradiction; OP-48c explicitly engaged and strengthened (residue shrunk from 'GPT formalization choice' to 'normal-state postulate + composition-primitive choice'); OP-49 near-no-go untouched; HYP-ENCODING-SCREEN reinforced (the state postulate exposed as an installation choice at yet another layer). The headline's 'forced' is upheld as forced-modulo-rejecting-the-question: the SS reading is inexpressible as a weakening of any reconstruction axiom and vacuous where expressible, but the forcing step remains the track's own INFERENCE, high, not a theorem — CONCLUSION §5 wording should preserve that register.

Net. RESOLVED-POSITIVE upheld with the corrections above binding. The iteration's product — the LT-sep theorem-let and the (PS)/(LT-sep)/(LT-dim) decomposition — is correct, new, and sharpens horn C beyond the mission conditional. OP-50 remains CONTESTED, now genuinely three-cornered. Object-level: still NOT-YET-PHYSICS; no distinguishing experimental channel emerged from this track.