Technical note BMR-1 of the RCD program (Radial Coherential Dynamics), continuing "Relational Memory under Causal Erasure" (10.5281/zenodo.23630978). The previous record found an exact memory-flow law on the star family, which carries a single memory mode. This note asks for the minimal causal architecture able to carry two independent memories and survive its own coarse-graining. Three exact results. (1) Branch–rank lemma: if the N erasure positions produce r distinct branches, rank M_N ≤ r − 1, with equality under affine independence; two modes require at least three causally inequivalent erasures. (2) The P2_n sector (a chain with one pendant element at height ≤ 2) is closed under erasure for all n ≥ 4, has unbounded height and two exact memory modes: exact transfer map R_n, two flow invariants (y, J), closed-form solution, Hermitian memory M_n with rank 2 iff uv ≠ 0, and linear transport of (|u|², Re ūv, Im ūv, |v|²) with spectrum {a², ac, ac, c²}, a = (n−2)/n, c = (n−3)/n; the relative phase is carried as an independent memory component. (3) Step-closure theorem: the support of every Rideout–Sorkin (RS) amplitude is fixed by the local pair (|↓x|, #max ↓x) and always contains the antichain, so no closed family omitting it, P2_n in particular, admits an RS realization; the star is the only RS-supportable family found. An exact classification of RS closure families 5→4 on restricted supports finds only the star line, plus a certified isolated conjugate pair that closes one step and not the next. The obstruction is against the locality of the RS amplitude support, not against scalar amplitudes as such. Includes the erasure automaton of unlabeled posets up to N = 7 (1, 2, 5, 16, 63, 318, 2045 classes) and all scripts with a SHA-256 manifest. Exact results are stated as such; the enumeration of isolated points on large supports is not certified. English and Spanish versions; the Spanish text prevails. Developed at the Panel Cerezo: ChatGPT (formulation of BMR-1, branch–rank lemma, Hermitian correction), Claude (automaton, P2_n, step-closure theorem, classification, verification and drafting), under the direction and adjudication of the principal.
Technical note BMR-1 of the RCD program (Radial Coherential Dynamics), continuing "Relational Memory under Causal Erasure" (10.5281/zenodo.23630978). The previous record found an exact memory-flow law on the star family, which carries a single memory mode. This note asks for the minimal causal architecture able to carry two independent memories and survive its own coarse-graining. Three exact results. (1) Branch–rank lemma: if the N erasure positions produce r distinct branches, rank M_N ≤ r − 1, with equality under affine independence; two modes require at least three causally inequivalent erasures. (2) The P2_n sector (a chain with one pendant element at height ≤ 2) is closed under erasure for all n ≥ 4, has unbounded height and two exact memory modes: exact transfer map R_n, two flow invariants (y, J), closed-form solution, Hermitian memory M_n with rank 2 iff uv ≠ 0, and linear transport of (|u|², Re ūv, Im ūv, |v|²) with spectrum {a², ac, ac, c²}, a = (n−2)/n, c = (n−3)/n; the relative phase is carried as an independent memory component. (3) Step-closure theorem: the support of every Rideout–Sorkin (RS) amplitude is fixed by the local pair (|↓x|, #max ↓x) and always contains the antichain, so no closed family omitting it, P2_n in particular, admits an RS realization; the star is the only RS-supportable family found. An exact classification of RS closure families 5→4 on restricted supports finds only the star line, plus a certified isolated conjugate pair that closes one step and not the next. The obstruction is against the locality of the RS amplitude support, not against scalar amplitudes as such. Includes the erasure automaton of unlabeled posets up to N = 7 (1, 2, 5, 16, 63, 318, 2045 classes) and all scripts with a SHA-256 manifest. Exact results are stated as such; the enumeration of isolated points on large supports is not certified. English and Spanish versions; the Spanish text prevails. Developed at the Panel Cerezo: ChatGPT (formulation of BMR-1, branch–rank lemma, Hermitian correction), Claude (automaton, P2_n, step-closure theorem, classification, verification and drafting), under the direction and adjudication of the principal.
We present a development of the RCD program (Radial Coherential Dynamics) in the complex scalar sector of quantum sequential growth of causal sets. The central question is not only how a causal history grows, but how much of the information about HOW IT WAS FORMED can survive when one of its elements is erased without privileging any temporal label. The central result is an exact separation between coherent inheritance and memory of the cut. For the erasure branches a_i (i = 0, …, N−1), define the branch matrix 𝒜_N = N^(−1/2) [a_0, …, a_{N−1}], the uniform register mode |u_N⟩ = N^(−1/2)(1, …, 1), and Q_N = I − |u_N⟩⟨u_N|. The memory matrix M_N = D_mix − D_coh = 𝒜_N Q_N 𝒜_N† ⪰ 0 identifies exactly the memory sector orthogonal to coherent inheritance. Its trace, Tr M_N = (1/2N²) Σ_{i,j} ‖a_i − a_j‖², is the pairwise dispersion of the causally distinct erasure alternatives: relational memory is precisely that dispersion. The Relational Inheritance Bifurcation theorem (BHR-3) shows already at N = 3 that the inherited amplitude can coincide exactly while non-zero memory of the mode of formation persists: Δ₃^coh = 0 together with Ω₃ > 0, where D_mix − D_coh = Ω₃ · [[1, −1], [−1, 1]]. At the exact point r = 1, Ω₃ = 1/294; the reading with the register traced out has spectrum {147/149, 2/149} and visibility 145/149, while the coherent reading remains pure. One and the same observable inheritance can therefore coexist with residual relational information about its formation. The second main result is an exact memory-flow law. On the star family 𝒮_N — an antichain of N−1 elements with one element above all of them — covariant coherent erasure is closed for every N. With p = ε/(t₀+ε), p′ = (N−1) p / N, P := p/N = constant, and the complete coupling trajectory is ε_n = n ε_N / (N + (N−n) ε_N). The memory has rank one, M_N = (N−1)|P|² J with J = [[1, −1], [−1, 1]], so that Ξ := M_N/(N−1) = |P|² J is an exact invariant of the flow, M_{N−1} = M_N − Ξ, and 𝓜_N − 𝓜_{N−1} = 2|P|²: each coarse-graining step transports an exactly determined amount of relational memory. The star spine 𝒮₄ also appears as an exact component of the closure variety of the 4→3 flow. The non-degenerate regular branch C₁₂ is born tangent to this spine and inherits, at leading order, its memory-contraction factor 2/3. The deposit also contains the prior results leading to this structure: absence of a non-trivial scalar fixed point under uniform incoherent erasure (𝔗_mix); absence of a non-trivial fixed point under the uniform coherent superposition (𝔗_coh), which equals the classical PF-0 condition complexified; the need to formulate erasure covariantly by classes; the rational renormalization map on the closure variety; and the minimal register compression, rank D_mix = rank M_N + 1. These results motivate the Principle of Relational Memory of the Cut, dynamical version (PMRC-D): every fundamental physical differentiation that produces causally distinguishable histories must also determine, through a covariant rule, which information of that differentiation remains recorded and which information may interfere coherently. The emerging working thesis is: "The inheritance of a history does not by itself determine the memory of how it was formed." The program remains open toward closed sectors with several independent memory modes and a covariant dynamics capable of transporting those modes under coarse-graining. Exact results are N ≤ 4 except those with closed proofs for all N (the star spine and its flow); the 5→4 step outside 𝒮_N is numerical; the physical universality of PMRC-D remains open. No claim is made about the emergence of time, Mermin's 25 % or the necessity of H > 1 as a law. Deposit contents: RCD Adjudication Record v0.2 (signed Spanish original and English version; the Spanish text prevails), diligence notes 𝔗_mix-0, 𝔗_coh-0/𝔗_flow-0 and BMR-0 (Relational Memory), reproducible code with SHA-256 manifests, and README. Panel Cerezo credits: ChatGPT (formulation of the diligences, BHR-3, PMRC-D, 𝒜QA† form, invariant Ξ, series of C₁₂) and Claude (exact computation, lemmas, curve C and map R, star spine 𝒮_N, cross-verification and drafting), under the direction and adjudication of the principal.
Technical Note JANO-BR-0 of the RCD program (continues “La doble herencia de la coherencia”, DOI 10.5281/zenodo.23520439; the Adjudication Record “Relational Memory under Causal Erasure”, DOI 10.5281/zenodo.23717352; BMR-1, DOI 10.5281/zenodo.23717139; and REP-0, DOI 10.5281/zenodo.23774093). Question: can the heritage/memory bifurcation of BHR-BMR become a physical readout of one and the same circuit? Results (exact, no hardware): in BHR-3 the mixed relational-memory state ρ_mix(r) has rank 2 whenever Ω₃ > 0, so a single qubit c purifies it, |Ψ(r)⟩ = √λ₊|0⟩|ψ₊⟩ + √λ₋|1⟩|ψ₋⟩, with a three-gate {Ry, CNOT} construction and a closed map r ↦ (θ_c, η); tracing out c returns ρ_mix (R⊥), while measuring c in the HJW basis leaves the system exactly in the coherent-heritage state ρ_coh with probability p_coh = Tr D_coh/Tr D_mix and in the residual-memory state |−⟩ (independent of r) with probability p_M; the three readouts are tied by the exact identity V_mix = p_coh·V_coh − p_M (at r = 1: 145/149 = 147/149 − 2/149). Verified against the original JANO-1 code: the five-qubit JANO-1 circuit with α = α_BR(r) = arcsin √λ₋(r) is literally this purification, and its frozen observables reduce to P_rec = λ₊ and R∥ = R⊥ = √(2λ₊λ₋) = √S_lin(ρ_mix) — a spectral readout of the memory — while the HJW readout is the complementary face that separates heritage from memory. Stated limit: the construction reproduces ρ_mix(r) by design and is not a falsifiable prediction of RCD; hardware stays closed. Scripts with SHA-256 manifest. English and Spanish versions; the Spanish text prevails.
Technical Note CC (CC-0 to CC-3) of the RCD program (continues I-F0, DOI 10.5281/zenodo.24064727). Question: can a single finite coherence ρ ≥ 0 on ℂ^N, with no prior spacetime, produce a four-dimensional scenario with signature 1+3 and a metric g? This note records a construction that reached an explicit metric and was closed by a pre-registered death condition; it does not claim a derived spacetime. Prior art comes first. A two-report prior-art audit with cross-verification (reports, log and consolidation included) classifies as INHERITED: the cone of 2×2 Hermitian matrices with 4·det as the Lorentz form and SL(2,ℂ) (Baez; Arrighi–Patricot), its physical reading in the qubit — velocity, mass, purity, filtering as Lorentz, concurrence, a pure state being "like a photon" (Chen et al.; Bogdanov et al.; Burns–Greenfield–Dressel; Verstraete et al.; Rungta et al.; Fullwood–Vedral–Guzmán-González) —, the set of compressions (Fan–Pall 1957), the Löwner order as causal order (Mori–Šemrl), the density of compressions as a projection of an orbital measure (Olshanski; Faraut; Baryshnikov), the logic "causal order + volume = metric" (Finkelstein 1969; Hawking–King–McCarthy; Malament; Bombelli–Lee–Meyer–Sorkin) and the program of obtaining 3+1 from binary alternatives (Weizsäcker; Finkelstein; Müller–Masanes; Höhn–Müller; Gil). Three close neighbours (Finkelstein 1969, Brody–Hughston 2005, Gil 2026) are cited as the audit reports them, with a declaration of how each was read. OWN, as far as the two searches reached (medium confidence): (1) reading the views V†ρV of an N-level coherence through framed binary windows V ∈ St(2,N) as events — the spectral causal body 𝒟(ρ) = {2λ_{N−1} ≤ u ≤ 2λ₁, 2λ_N ≤ v ≤ 2λ₂, u ≥ v}, a 4-dimensional region bounded by null surfaces at the eigenvalues of ρ, with the trace as time function; (2) the soldering on the space of windows, e_P(X) = X†B + B†X with B = P⊥ρP: signature (1,3) plus a kernel of dimension 4N − 12 exactly where rank B = 2, and no timelike directions for pure ρ; (3) the coherential variance M̄(a) = E[V†ρ²V | a] − a² in closed form (M̄ = (α − a)(a − β) in the central chamber); (4) a specific metric: the pushed-forward invariant measure of the views as volume (Principle of Coherential Volume) and the cone by pullback of the coherential diffusion tensor (Rule C-0); (5) the computed results. Four postulates are declared, not derived: the governing rule C ⟹ g_μν (thesis of the program, not a result), Gate F (a framed binary cut is physical data), the Principle of Volume and Rule C-0. Results. CC-1 (N = 4): closed-form density of views, constant in the core; the volume metric g = (w/8)^{1/2}η has a flat core, curved outer chambers, R ~ d^{−3/2} at the boundary and undeformed cones; a 200,000-sample check agrees (one badly designed criterion failed, is archived, and was rerun with a corrected, pre-fixed criterion). CC-2 (N = 4): with Rule C-0 the metric is no longer conformally flat — Weyl² ≠ 0 in the four chambers and non-radial null geodesics are deflected — but at the outer spectral boundaries R ~ d⁻¹ and Kretschmann ~ d⁻³. CC-3 (N = 5, new input formulas checked by sampling): same qualitative picture, with R ~ d^{−5/2} and Kretschmann ~ d⁻⁵, i.e. stronger than for N = 4. The three death conditions (isotropic variance everywhere; curvature divergence at the outer boundaries; no deflection of null geodesics) and their operational definitions were written in the script headers and hashed before running; the second one triggered in both computations and the principal ratified both closures. One spectrum per N was computed; N ≥ 6, internal chamber walls and degenerate spectra were not. Not claimed: spacetime, gravity, Einstein's equations, absence of singularities. Scripts, outputs and SHA-256 manifest included; all charges of the adversarial rounds documented. English and Spanish versions; the Spanish text prevails.
Technical Note I-C of the RCD program, Act I (continues I-B1, DOI 10.5281/zenodo.23990720; I-B0, DOI 10.5281/zenodo.23965760; and I-A0/I-A1, DOI 10.5281/zenodo.23961107). I-B1 left the boundary that every realization is a sample entering conditionally; I-C asks whether an event can be registered without consuming the coherence needed for the next one, leave an inheritable trace, and take part in an inheritance network. In the tracial sector ρ = P/r (r = 2, 3, 4; no hardware; theorems with numerical control, 92 checks with declared tolerances, all passed), results: (I) Register-with-Residual-Capacity Theorem. A rank-one reading on the whole system admits only nuclear instruments (Haapasalo–Pellonpää) and hence breaks entanglement (Horodecki–Shor–Ruskai); the family A_η(Q) = (1−η)P + ηrQ with Lüders operation L_Q = aP + bQ gives, for 0 < η < 1, an informative reading (I_η > 0, with I₁ = ln r − H_r + 1) together with a non-entanglement-breaking depolarizing channel, p_η = 1 − b²/(r+1) > 1/(r+1), whose posterior state (1−η)P/r + ηQ is the I-B1 segment produced by a mechanism. Exact non-Markovianity on labels, k(Q₃|Q₁,Q₂) − k(Q₃|Q₂) = η²√(1−η²) in r = 2. Finite-history reversal theorem: w_n(Q₁…Q_n) = w_n(Q_n…Q₁) for every n, so no arrow comes from label statistics. Four separate persistences: the classical archive; the averaged residual, decaying as ηp_ηⁿ and never zero; entanglement with discarded registers, dying at n_EB = ⌈ln(1/(r+1))/ln p_η⌉; and entanglement with kept registers, with Schmidt rank r along every trajectory and the exact bound λ_min ≥ [(1−η)/(1+(r−1)η)]ⁿ/r — flagged-survival corollary: with the classical register kept the negativity is an integral of branch negativities and stays positive for every finite n; the classical register is the flag that preserves the resolution of entangled branches, not a source of entanglement. (II) Inheritance network, ratified conditionally under a declared connection scheme H_con (carrier registers): acyclicity by strict inclusion, invariance of density, state, registers and predecessors under all compatible linear extensions and relabeling, rhombus with incomparability sustained by commutation and no-signaling, and the exact counterexample 13/8, 5/4, 1/16 showing that a product label does not imply independence (local statistics factorize iff the posterior is a product). Orientation: the boundary source dies (P/r has a preimage; in a qubit L_{Q⊥}L_Q = √(1−η²)P), the contraction source belongs to the average, and accumulation remains declared. Locality no-go: the Kraus operators span M_r, the commutant is trivial and Fix(Φ_η) = ℂP, even for product labels on ℂ²⊗ℂ²: locality cannot be extracted from the algebra of the global instrument. Diagnostics: Rideout–Sorkin Bell causality transcribed to the label law fails; the prefixed rhombus network is not sprinkling-like. Not claimed: time, space, gravity, irreversibility, Bell causality, asymptotic quantum capacity, or the derivation of η, accumulation, connections or factorization; the event debt (No Free Actualization) remains open. Scripts with SHA-256 manifest; run incidents documented. English and Spanish versions; the Spanish text prevails.
Technical Note I-B1 of the RCD program, Act I (continues I-B0, DOI 10.5281/zenodo.23965761, and I-A0/I-A1, DOI 10.5281/zenodo.23961108). I-B0 derived, from the frozen input of Act I (an algebra M_d, a state ρ and its full symmetry group G₀), the space 𝒪_ρ of elementary distinctions, its canonical law μ_ρ, the register interface E_ρ and the channel 𝓜_ρ. I-B1 asks whether that structure can produce a concrete record — operationally a pure state δ_Q of the classical algebra C(𝒪_ρ) — without selection, clock, asymmetry or primitive randomness. Result (theorems with numerical control on 17 states in d = 2, 3, 4; no hardware): the No-Free-Actualization Theorem, with four distinct and complementary obstructions. (1) No deterministic G₀-equivariant rule takes μ_ρ to any δ_Q when rank ρ > 1; a concrete record reduces the state's stabilizer from G₀ to H_Q. (2) Every instrument compatible with E_ρ is measure-and-prepare (Haapasalo–Pellonpää), the covariant ones correspond to stabilizer-invariant states, and the segment tQ + (1−t)ρ is covariant with identical statistics: POVM plus symmetry do not fix the post-measurement update; the instrument preparing Q is selected by a declared consistency condition (H_PB, Petz–barycenter), legitimized by the unique affine weak-* continuous extension of the Petz recovery. (3) The canonical closure T_ρ = 𝓜_ρ∘𝓑_ρ is a smoothing kernel (T_ρν ≪ μ_ρ for every ν, never a delta), positive (0 ≤ T_ρ ≤ I) and reversible (detailed balance) with respect to μ_ρ; for ρ = P/r its spectrum is {1, 1/(r+1) with multiplicity r² − 1, 0} with exact contraction (1/(r+1))^{n−1} after the first step and D(Q‖𝓑_ρ(T_ρδ_Q)) = ln((r+1)/2); in d = 2 the spectrum is {1, √(λ₁λ₂) double, 0}, jumping from 1/2 to 1/3 at degeneracy; and a coalescence theorem shows μ_ρ and its second moment discontinuous at every coalescence of positive eigenvalues (Var Tr(P₁Q) jumps from 0 to λ²n₁n₂/(n₁+n₂+1)); every possible distinction lies at the same distance from the barycenter, ‖Q − ρ‖² = 1 − Tr ρ². (4) μ_ρ is realized as Q = gQ₀g* with g Haar-distributed on G₀: the structure derives the law of the sample, not the sample. Stated boundary: it is not claimed that no extension of RCD can derive the event, nor that the event is a fundamental primitive; μ_ρ ⇝ δ_Q is recorded as an open debt; I-A/I-B derive new structure from the Act I input but do not derive that input from a structureless coherence (an upstream question). For I-C the event enters only as a conditional input, and the next wall is stated: can an event be registered without consuming the coherence needed for the next one? Includes an erratum for the two earlier deposits (17 control states, not 16; 14 of rank > 1, not 13). Scripts with SHA-256 manifest; the original run code with a corrected spectral-control error is kept for traceability. English and Spanish versions; the Spanish text prevails. "Erratum (v0.2, 5 November 2026): v0.1 of this note and v0.2 of the I-A1 report stated 16 control states; there are 17 (3 in d = 2, 5 in d = 3, 9 in d = 4; 3 pure and 14 of rank > 1), all contained in ia1.out and all passing the gates. Documentary error with no effect on any result or theorem; the I-A0 count (13 cases) is correct. This version corrects the note and the I-A1 report; code and outputs are unchanged."
Technical Note I-B0 of the RCD program, Act I (continues I-A0/I-A1, DOI 10.5281/zenodo.23961108, and MEM-0, DOI 10.5281/zenodo.23926847). Act I left wall I-A resolved in the finite species: the pair (M_d, ρ) canonically determines the orbit 𝒪_ρ of elementary distinctions (rank-one projections) whose Haar barycenter is ρ, selecting none of them. I-B0 asks whether those same data generate, on their own, a register interface — a covariant quantum-to-classical channel whose outcome space is not added from outside — that loses the recoverability of every concrete distinction, with no clock, no preferred basis and no external cut. Results (theorems with numerical control in d = 2, 3, 4 on sixteen states; no hardware): (i) Canonical Register Interface Theorem — E_ρ(dQ) = ρ^{-1/2} Q ρ^{-1/2} dμ_ρ(Q) is a G₀-covariant POVM on 𝒪_ρ integrating to the support (the square-root / pretty-good measurement of the canonical ensemble); the channel 𝓜_ρ(σ)(dQ) = Tr(σ E_ρ(dQ)) satisfies 𝓜_ρ(ρ) = μ_ρ; and its Petz recovery map with respect to ρ is exactly the barycentric preparation ν ↦ ∫Q dν of I-A1, so that the I-A1 barycenter and the I-B0 Petz recovery are the same object seen from both sides; the chain (M_d, ρ) → 𝒪_ρ → μ_ρ → E_ρ → 𝓜_ρ requires no new hypothesis; canonicity is self-aligned uniqueness: among covariant rank-one-seed POVMs on 𝒪_ρ, E_ρ is the only one whose response to Q peaks at Q itself (the others are its central relabelings and, only for ρ = P/2, an anti-aligned measurement exhibited in the review round). (ii) Petz Lock — D(Q‖ρ) = S(ρ) on the whole orbit; the gap Δ_rec = D(Q‖ρ) − D(𝓜_ρ(Q)‖μ_ρ) = χ − I(E_ρ) is Q-independent and vanishes iff rank ρ = 1: for rank greater than one the channel returns ρ but is not sufficient for any pair (Q, ρ), i.e. it does not fully preserve any concrete distinction relative to the reference coherence (not absolute irrecoverability). (iii) Exact benchmark Δ_rec(P/r) = H_r − 1 (harmonic number), equal to the Holevo gap of the accessible information of the uniform ensemble (Jozsa–Robb–Wootters); closed form in d = 2, Δ_rec = S(ρ) − ln((1+s)/2) − 1 + s with s = |λ₁ − λ₂|; and a discontinuity of the gap at spectral degeneracy, 2 ln 2 − 1 ≈ 0.386 versus 1/2 at ρ = I/2, because the stabilizer and the canonical orbit change. Stated boundary: no concrete record is adjudicated — 𝓜_ρ(ρ) = μ_ρ, not δ_Q, and no deterministic equivariant rule can select δ_Q; stochastic selection remains open; the realization of an outcome is the question of I-B1, not opened here. Declared heritage: pretty-good measurement, Petz recovery, Barnum–Knill, Holevo's covariant measurements; what is RCD's own is the input, since the ensemble is derived from symmetry rather than postulated. Scripts with SHA-256 manifest. English and Spanish versions; the Spanish text prevails. Erratum (v0.2, 5 November 2026): v0.1 of this note and v0.2 of the I-B0 report said “sixteen states” and “thirteen of rank > 1”; there are 17 control states (3 pure and 14 of rank > 1), all contained in ib0.out and all passing the gates. Documentary error with no effect on any result or theorem. This version corrects the note and the I-B0 report; code and output are unchanged.