Path-Dependent Collapse from Higher-Categorical Obstructions: A Geometric Framework Beyond Standard Quantum Mechanics with Application to Nuclear Fission

Path-Dependent Collapse from Higher-Categorical Obstructions: A Geometric Framework Beyond Standard Quantum Mechanics with Application to Nuclear Fission Andrei T. Patrascu 1 1 FAST Foundation, Destin FL, 32541, USA email: andrei.patr[email protected] Collapse in quantum mechanics is traditionally modeled as entanglement with an environment, producing decoherence once environmental degrees of freedom are traced out. While this explains the loss of coherence, it does not intrinsically predict the basis, locus, or timing of collapse, nor does it account for protocol dependence. We propose that collapse originates in higher-categorical coherence breakdown (HCCB): structural obstructions in the geometry of moving projector bundles associated with relevant subspaces. Using the quantum metric and curvature of these projectors, we derive a collapse functional whose integrated value determines when collapse occurs. Central (triangle-level) anomalies yield only phases, while non-central (HCCB) obstructions produce Lindblad jump operators, making collapse a structural inevitability rather than an ad hoc postulate. This framework predicts path-dependent collapse: two control trajectories with the same endpoints but different order or shape can yield different coherence loss, a signature not explained by standard decoherence models. As an example, we apply this approach to nuclear fission, where the relevant subspaces are left and right fragments defined by the evolving neck. HCCB predicts not only the fragment-number basis of collapse and its locus at the neck, but also order-dependent differences in the yields and widths of mass distributions—an added predictive layer beyond shell corrections and environment-induced decoherence. Beyond nuclear physics, the same geometric formalism provides new control knobs for quantum processors: minimizing collapse during gates, accelerating collapse for measurement, and diagnosing geometric collapse via order-dependent protocols. I. INTRODUCTION A. Motivation Context. In the textbook formulation of quantum mechanics, measurement is modeled by an abrupt, stochastic collapse of the state vector, typically introduced axiomatically (Born’s rule, von Neumann’s projection postulate) [ 1 , 2 ]. Modern open–quantum–systems theory replaces this postulate with an effective dynamical account: system–environment entanglement followed by tracing out environmental degrees of freedom yields reduced dynamics described (under broad conditions) by a Markovian Gorini– Kossakowski–Sudarshan–Lindblad (GKSL) master equation [ 6 , 25 , 26 , 94 ]. Decoherence theory then explains why interference terms vanish in certain “pointer” bases (einselection) and how classical records emerge [ 29 , 77 , 78 ]. Trajectory and continuous–measurement formalisms further refine this picture at the level of single realizations [102, 103]. What remains unexplained. While these frameworks account for how coherence is suppressed, they do not, by themselves, predict three structural features of collapse: (i) the basis in which records stabilize (typically posited or inferred a posteriori); (ii) the locus (where along a controlled evolution the state becomes effectively classical); and (iii) the timing (when collapse happens for a given protocol). In practice, basis choice is often tied to the observable that couples most strongly to the bath; the locus is asserted to be “at measurement”; and timing is identified with a fitted T2 or a bath–set rate [ 78 , 94 ]. None of these are intrinsic predictions of standard theory; they are model inputs or phenomenological outcomes. Moreover, in conventional treatments (fixed endpoints and stationary noise), the path by which a Hamiltonian is reached typically has no leading–order effect on the accumulated decoherence, aside from familiar Landau–Zener or nonadiabatic corrections [85]. A geometric/categorical perspective. We advance a complementary point of view: collapse is a structural effect of how the relevant subspace of the system moves under external controls. Let q = ( q1, . . . , qd ) denote control parameters (“knobs”), and let P ( q )be the orthogonal projector onto the subspace carrying the degrees of freedom that are operationally relevant (e.g., a dressed qubit, an emergent fragment subspace in nuclear dynamics). The map q7→ P ( q )endows the parameter manifold M with a natural projector bundle geometry: gij(q) = Tr ∂iP ∂jP, Fij(q) = P[∂iP, ∂jP]P, (1) 2 namely the pullback of the Fubini–Study metric and its associated connection/curvature [ 17 , 96 – 100 ]. This bundle–theoretic structure distinguishes two kinds of “noncommutativity” of control operations: ∂iP= [Ki, P] + Ei, Ei:= Q ∂iP P, (2) with Q = 1 −P and K† i = Ki . The central (“triangle–level”) part [ Ki, P ]generates motions within the relevant subspace, leading to geometric phases and holonomies (Berry/Wilczek–Zee) but no leakage [ 97 , 98 ]. The non–central part Ei maps between P and its complement Q and therefore represents a genuine coherence–breaking channel. Higher–categorical coherence breakdown (HCCB). We use the term higher–categorical coherence breakdown (HCCB) to denote precisely the situation Ei6 = 0 on a set of nonzero measure along a control path. In the time–dependent projection formalism (Nakajima–Zwanzig/Davies–type reductions), Ei is the seed of completely positive (CP) jump operators Li∝Q ∂iP P in the reduced dynamics on Ran P [ 91 – 94 ]. As a result, the instantaneous dephasing (collapse) rate of coherences in the P –basis becomes a geometric functional of the path velocity ˙q: γ(t) = η˙qi˙qjgijq(t),Γ[q(·)] = ZT 0 γ(t)dt, (3) with a scale factor η fixed by the relaxation rate(s) in the fast complement Q (e.g., Purcell loss in cQED). Equation (3) exhibits two consequences that are not predicted by standard endpoint–only or stationary– bath models: (i) a threshold law for collapse timing, Rt? t0˙qi˙qjgij dt ≃ Θ; and (ii) path dependence: for two protocols with identical endpoints and duration, the accumulated exponent Γcan differ when the curvature two–form Fis nonzero on the region enclosed by the paths (order–of–operations matters). Physical meaning. Intuitively, gij measures how far the relevant subspace “tilts” in Hilbert space per unit change of controls, so k˙qk2 g = ˙qi˙qjgij is the speed squared with which the projector is dragged. The non–central derivative Ei is the precise leakage operator that transfers amplitude out of P , and its norm is governed by g . Central curvature (pure phases) corresponds to commuting diagrams up to a scalar and does not produce collapse; non–central curvature corresponds to higher–coherence failure (diagrammatic noncommutativity not absorbed by a phase) and does produce a CP dissipator [ 94 , 97 , 98 ]. In this sense, the metric/curvature pair ( g, F )are the geometric shadows of deeper categorical commutativity conditions: when only central anomalies are present, one observes phases; when higher (non–central) obstructions appear, one observes bona fide collapse. Why this matters empirically. Because γ depends on the trajectory q ( t )via g and F , this framework predicts protocol–dependent collapse: two sequences of control operations with the same endpoints and total time (e.g., amplitude–then–detuning vs detuning–then–amplitude in a driven qubit; elongation– then–neck vs neck–then–elongation in nuclear shapes) can yield measurably different coherence loss. This is a falsifiable signature beyond conventional decoherence modeling and, crucially, it provides new design handles: one can shape when collapse occurs (timing via speed), where it concentrates (locus at geometric “hot spots” where g spikes), and which degrees of freedom become classical records (basis fixed by the block structure of P ). As an emblematic example developed later in the paper, we show how this perspective explains the emergence of fragment–identity records (and their timing and locus) in nuclear fission while predicting path–dependent differences in fragment–mass distributions that go beyond standard liquid–drop plus shell–correction descriptions [107–109]. B. Problem Statement of the problem. In the standard open–quantum–systems account of measurement, one models the system S coupled to an environment E , prepares an initial product state ρS (0) ⊗ρE , and evolves unitarily under USE(t) = Texp{−i ~Rt 0HSE(τ)dτ}, yielding the reduced state ρS(t) = TrEhUSE(t)ρS(0)⊗ρEU† SE(t)i.(4) Under broad Markovian and weak–coupling assumptions, ρS ( t )obeys a Gorini–Kossakowski–Sudarshan– Lindblad (GKSL) master equation [ 25 , 26 , 94 ]. This formalism compellingly explains how off–diagonal elements (coherences) in certain representations decay, i.e., decoherence. Yet, three structural features of “collapse” remain unpredicted by standard theory and are typically supplied by additional assumptions: 3 •Basis (preferred–basis problem). In which basis do classical records stabilize? •Locus. Where along a controlled evolution does the state become effectively classical? •Timing. When (under a given protocol) does collapse occur? Why the basis is not intrinsically predicted. Decoherence arguments often begin by positing a measurement–like coupling HSE = Pa|aiha| ⊗ Ba so that, for pure dephasing, the reduced density matrix in the {|ai} basis evolves as ha|ρS(t)|bi=ha|ρS(0)|biDab(t), Dab(t) = TrEe−i(Ba−Bb)t/~ρE,(5) and |ai are then identified with pointer states [ 29 , 77 , 78 ]. But unless one assumes this form of HSE (or a specific structure of Lindblad operators), the preferred basis is not determined a priori. In realistic, time–dependent settings, different system–bath couplings or unravelings (e.g., photon counting vs. homodyne detection) select different pointer bases even for the same GKSL generator [ 102 – 104 ]. The “predictability sieve” program ranks states by robustness/entropy production, but generally yields afamily of nearly equivalent candidates rather than a unique basis [ 34 , 77 ]. Alternative foundational frameworks (consistent/decoherent histories [ 35 – 38 ], modal interpretations, or objective–collapse models [ 39 – 41 ]) change the postulates but still do not derive the pointer basis from control geometry of the experiment. Why the locus is not intrinsically predicted. In the open–systems picture, decoherence occurs continuously and often rapidly on a model–dependent time scale Tφ set by bath spectral densities [ 42 , 94 ]. One may declare the state “effectively classical” once coherences fall below some threshold, but the place along a driven protocol at which that threshold is crossed is then an extrinsic criterion. In consistent histories, decoherence is diagnosed by the vanishing of off–diagonal terms in the decoherence functional for a given coarse–graining [ 35 – 37 ], but the choice of coarse–graining fixes the locus rather than predicting it. In continuous–measurement theory, the locus is tied to the measurement record and its bandwidth [ 103 , 104 ]; again, it is a property of the monitoring scheme, not an intrinsic feature of the system’s controlled motion through Hilbert space. Why the timing is not intrinsically predicted. Standard theory provides decay rates (e.g., T1 , T2 ) from microscopic bath couplings, but, given a time–dependent Hamiltonian HS ( q ( t )) that implements a target unitary between the same endpoints in the same total time, there is typically no structural reason in the Markovian limit for two different orderings of controls (path A vs. path B ) to yield different integrated decoherence, beyond familiar nonadiabatic transitions (Landau–Zener physics) [ 43 , 85 ] or explicit time–dependence of the rates. In other words, timing is set by fitted or computed rates, not by a threshold law that depends on intrinsic geometry of the controlled subspace. Open–system geometric phases do exist [95], but they quantify phase holonomy, not a path–dependent collapse functional. Summary of the gap. To summarize: (i) the basis of collapse is tied to assumed couplings or monitoring, not derived from structural features of how the relevant subspace moves; (ii) the locus is prescribed by external thresholds, coarse–grainings, or measurement bandwidths; (iii) the timing follows bath–set rates rather than a path–integrated, geometry–driven criterion. Even objective–collapse theories (GRW/CSL) replace environmental decoherence with stochastic terms but still postulate collapse dynamics rather than derive basis–locus–timing from control geometry [ 39 , 41 ]. This paper addresses the gap by introducing a geometric/categorical mechanism—higher–categorical coherence breakdown—that computes basis, locus, and timing from the projector bundle q7→ P ( q ), and predicts a distinct, falsifiable signature: path–dependent collapse for protocols with identical endpoints and duration. C. Proposal Core idea. We propose that quantum–state collapse originates in higher–categorical coherence breakdown (HCCB): a structural obstruction that arises when the relevant subspace of a quantum system, represented by an orthogonal projector P ( q )depending smoothly on external controls q = ( q1, . . . , qd ), is moved through Hilbert space in a way that cannot be absorbed by inner (central) unitary rotations within Ran P . Concretely, the map q7→ P ( q )defines a vector bundle over the parameter manifold M whose differential geometry—encoded by the quantum geometric tensor (QGT) and its decomposition into a real (metric) and imaginary (curvature) part—governs the system’s response [ 51 , 96 – 101 ]. When the differential of Pdecomposes as ∂iP= [Ki, P] + Ei, Ei:= Q ∂iP P, Q := 1 −P, (6) 4 the commutator term [ Ki, P ]generates central (triangle–level) motions that produce phases/holonomies but no leakage out of Ran P [ 63 , 97 , 98 ]. By contrast, the non–central block Ei maps between Ran P and its complement and thus constitutes a genuine obstruction to coherence persistence within P . We identify HCCB with the occurrence of Ei6= 0 along a control path on a set of nonzero measure. Geometric data and physical meaning. The projector bundle carries the pullback of the Fubini–Study structure, with quantum metric and (nonabelian) curvature gij(q) = Tr ∂iP ∂jP, Fij(q) = P[∂iP, ∂jP]P, (7) which may be equivalently viewed as the real and imaginary parts of the QGT restricted to the relevant bundle [ 96 , 99 , 101 ]. Physically, gij quantifies the speed–squared with which the relevant subspace tilts in Hilbert space under dqi , while Fij quantifies the noncommutativity of sequential control moves ( i then j vs. j then i ). When the relevant subspace arises as a spectral cluster of an instantaneous Hamiltonian H(q), one can represent P(q)by a resolvent integral, P(q) = −1 2πi IC (z−H(q))−1dz, ∂iP=−1 2πi IC (z−H)−1(∂iH)(z−H)−1dz, (8) so that Ei is controlled by off–diagonal matrix elements of ∂iH across the P / Q energy gap [ 55 , 74 , 85 , 86, 113]. Reduced dynamics from time–dependent projections. Consider the exact unitary evolution on H driven by H ( q ( t )) and partition the state ρ into P/Q blocks. In the time–dependent projection–operator formalism (Nakajima–Zwanzig–Davies), under a standard separation of time scales where the Q sector relaxes rapidly (or is efficiently drained by a bath), the reduced state ρP := PρP obeys a completely positive (CP), time–local master equation to second order in velocities ˙q[62, 91–94]: ˙ρP=−i ~[Heff(t), ρP] + X i,j ˙qi˙qjLiρPL† j−1 2{L† jLi, ρP}+O( ˙q3),(9) with jump operators proportional to the non–central derivatives Li∝Ei=Q ∂iP P, (10) and an Heff that contains the projected Hamiltonian and central geometric terms (adiabatic connection) [ 86 , 94 , 99 ]. Equation (9) shows that only the non–central part of the projector motion seeds CP dissipation within the relevant subspace; pure central (inner) motions generate phases but no collapse. This is the operational content of “coherence breakdown”: non–central geometric motion forces a dissipative channel in the reduced dynamics. Collapse functional and threshold law. The dissipator in (9) induces dephasing of off–diagonal elements in the P–basis at an instantaneous rate quadratic in the control velocity, γ(t) = η˙qi˙qjgijq(t),Γ[q(·)] = ZT 0 γ(t)dt, (11) where the prefactor η is set by relaxation rates in Q (e.g., Purcell loss in cQED, optical leakage in Λ–systems) or by engineered dissipation [ 74 , 75 ]. Equation (11) defines a collapse functional on the space of control trajectories. Collapse occurs once a geometric threshold is exceeded, Zt? t0 ˙qi˙qjgijq(t)dt &Θ,(12) with Θset by the initial coherence scale and the desired fidelity/visibility. Crucially, Γdepends on the path q ( · ), not merely on endpoints or the total time: for two protocols that trace the edges of a small rectangle in Min opposite orders (“A→B” vs. “B→A”), one finds ΓA−ΓB≈η TZZR∂q1g22 −∂q2g11dq1dq2,(13) an area law set by curvature of the metric tensor field (equivalently, by the nonabelian curvature F when expressed in a covariant form), providing a falsifiable signature of path–dependent collapse [99, 100]. 5 Predicting basis, locus, and timing. Within this framework, three longstanding questions are answered structurally: (i) the basis is the block structure of P ( q )—i.e., the labels that Ei and ∂iP actually mix/diagonalize (pointer states emerge as P –blocks stabilized against Q –leakage); (ii) the locus corresponds to regions of M where gij (or the operator norm of Ei )spikes (“geometric hot spots,” such as avoided crossings for dressed qubits or the neck surface in nuclear fission); (iii) the timing follows the threshold law (12) , i.e., it is set by the integrated metric speed along the realized path. These statements link the dynamics of collapse to differential geometry, rather than to ad hoc postulates or solely to stationary bath properties [62, 94]. Relation to, and distinction from, standard accounts. Decoherence theory attributes loss of coherence to system–environment entanglement, and continuous–measurement theory refines this at the trajectory level [ 68 , 102 – 104 ]. HCCB does not deny these mechanisms; instead, it identifies a structural source of collapse: the non–central geometry of moving projectors. In the absence of loss (lossless Q ), Ei6 = 0 still generates reversible leakage; with a lossy or rapidly relaxing Q , the same geometric seed yields an irreversible CP channel inside P . Thus environment sets irreversibility (the scale η ), whereas geometry sets propensity and selectivity (basis, locus, timing, and path dependence). This complementarity underlies both our foundational claims and the practical control protocols we develop later (e.g., minimizing Γ during gates or maximizing it for fast readout) [74, 75, 99]. Outlook. Sections that follow derive (9) – (129) explicitly, develop intuition via simple two–level examples (Bloch sphere), and present a worked application to nuclear fission where the relevant projectors track emergent fragment subspaces defined by the evolving neck. We show how HCCB predicts where collapse concentrates (neck surface), when it occurs (geometric threshold), and—importantly—how yields acquire path dependence for protocols with identical endpoints, a prediction that goes beyond conventional liquid–drop + shell–correction and environment–only accounts. D. Thesis: HCCB provides computable, structural predictions Statement. We claim that higher–categorical coherence breakdown (HCCB) renders collapse a computable and structural feature of driven quantum dynamics. Given a smooth map q7→ P ( q )from a control/parameter manifold M to orthogonal projectors on Hilbert space H , the geometry of the associated projector bundle—captured by the quantum metric gij ( q ) = Tr ( ∂iP ∂jP )and curvature Fij ( q ) = P [ ∂iP, ∂jP ] P —fully determines (i) the basis stabilized by collapse, (ii) the locus where collapse concentrates, (iii) the timing of collapse via a geometric threshold, and (iv) the protocol (path) dependence of collapse for control trajectories with identical endpoints. Crucially, these predictions arise from the non–central part Ei := Q ∂iP P of projector motion; central (inner) motions [ Ki, P ]yield only phases/holonomy and no dissipation [96–99]. Setup and reduced dynamics. Let q ( t ) ∈M be a differentiable control path and decompose ∂iP = [ Ki, P ] + Ei with Q = 1 −P . In the time–dependent projection framework (Nakajima–Zwanzig–Davies) with a fast, possibly lossy complement Q , the reduced state ρP = PρP obeys, to second order in velocities, a CP master equation ˙ρP=−i ~[Heff(t), ρP] + X i,j ˙qi˙qjLiρPL† j−1 2{L† jLi, ρP}+O( ˙q3), Li∝Ei,(14) with the instantaneous collapse (dephasing) rate quadratic in the metric speed, γ(t) = η˙qi˙qjgijq(t),Γ[q(·)] = ZT 0 γ(t)dt, (15) where η is set by relaxation in Q (e.g., Purcell rate in cQED, optical leakage in Λsystems) [ 74 , 75 , 94 ]. Equations (14)–(15) are the calculational backbone for the four claims below. (i) Basis: which subspaces are stabilized. Let P ( q ) = PµPµ ( q )be a resolution of the relevant projector into orthogonal blocks (candidate pointer sectors). Writing the off–diagonal coherence blocks ρµν =PµρPν(µ6=ν), one finds ˙ρµν(t) = −Γµν(t)ρµν(t) + skew–Hermitian terms,Γµν(t) = ˙qi˙qjΞ(ij) µν q(t)≥0,(16) with Ξ(ij) µν =hµ|∂iP ∂jP|µi+hν|∂iP ∂jP|νi−2 Re hµ|∂iP|νihν|∂jP|µi.(17) 6 The stabilized (pointer) basis is the one that diagonalizes the quadratic form Γ µν , i.e. the block structure for which off–diagonal decay rates are extremal/minimal. In spectral realizations (instantaneous eigenprojectors of H ( q )), Pµ are the dressed sectors and the above reduces to a Fubini–Study susceptibility projected onto block pairs [ 96 , 99 , 101 ]. Thus, HCCB derives the pointer basis from P ’s moving block structure, rather than assuming it from HSE [77, 78]. (ii) Locus: where collapse concentrates. Define the local collapse propensity Λ(q; ˙q) = max µ6=ν˙qi˙qjΞ(ij) µν (q),and k˙qk2 g= ˙qi˙qjgij(q).(18) Regions of M where gij (or Λ)spikes act as geometric hot spots in which leakage Ei is largest and collapse concentrates. In driven qubits, these are avoided–crossing neighborhoods or control points with large adiabatic gauge potentials [ 99 , 113 – 115 ]. In nuclear fission (treated later), hot spots occur at the neck surface, where the fragment–partition projector varies sharply with shape coordinates. (iii) Timing: when collapse happens. Integrating (15) gives a geometric threshold law: Zt? t0 ˙qi˙qjgijq(t)dt ≃Θ(ε),⇒ kρµν (t?)k ≤ εkρµν(t0)k,(19) with Θ( ε )set by the desired visibility/purity (a Grönwall–type bound). The timing t? therefore depends on the metric line–integral along the realized trajectory, not solely on bath parameters. This yields practical design handles: (a) dilate time to reduce Γ(speed control, cf. quantum speed limit analogies [83, 100]); (b) route through flatter regions of g(path control). (iv) Protocol dependence: how collapse depends on the path. For two protocols with identical endpoints and duration that traverse the boundary of a small rectangular region R ⊂ M in opposite orders (“A→B” vs. “B→A”), Taylor expanding gij and using Stokes’ theorem yields an area law: ΓA−ΓB≈η TZZR∂q1g22 −∂q2g11dq1dq2≡η TZZRKdS, (20) where K is a curvature–like scalar built from the metric (equivalently expressible via the nonabelian curvature F in a covariant gauge). Hence, order of operations matters for collapse even at fixed endpoints, a prediction beyond endpoint–only or stationary–bath models [ 95 , 99 ]. This path dependence is the key falsifiable signature of HCCB. Computability and experimental program. All quantities in (16) – (20) are either directly computable from device models (matrix elements of ∂iH across P/Q gaps via the resolvent formula [ 85 , 86 ]) or measurable tomographically (Ramsey visibility differences for A vs. B protocols; scaling with speed T−1 and with engineered relaxation η ) [ 102 – 104 ]. In this sense HCCB transforms collapse from a postulate/noise fit into a geometric design principle: it predicts the stabilized basis, locates collapse hot spots, sets timing by a threshold integral, and foresees protocol dependence testable with today’s platforms (superconducting circuits, trapped ions, NV centers) [74, 75, 112]. E. Contributions of this paper Overview. This work advances a geometric and higher–categorical account of quantum collapse and delivers four concrete, testable contributions: (i) a formal derivation of collapse from the non–central part of projector motion (curvature at the level of moving subspaces), (ii) an intuitive shadow picture that ties leakage amplitudes to measurable signals, (iii) a worked application to nuclear fission that predicts fragment–identity collapse (including the characteristic asymmetric A≈ 95 and A≈ 137 peaks after prompt emission) together with path dependence not derivable from standard QM models, and (iv) an outline of quantum computing applications that uses the same geometry to compile gates with reduced collapse and to accelerate measurement. (C1) Formal derivation of collapse from non–central projector curvature. We consider a relevant subspace represented by an orthogonal projector P ( q )depending smoothly on controls q = (q1, . . . , qd), with complement Q= 1 −P. The differential of Padmits the canonical decomposition ∂iP= [Ki, P] + Ei, Ei:= Q ∂iP P, (21) 7 where the inner (central) term [ Ki, P ]generates holonomy/phase within Ran P while the non–central term Ei maps between P and Q . Using time–dependent projection methods (Nakajima–Zwanzig–Davies) under a standard separation of time scales in which Q relaxes rapidly (or is lossy), we derive a completely positive, time–local reduced dynamics for ρP:= PρP, ˙ρP=−i ~[Heff(t), ρP] + X i,j ˙qi˙qjLiρPL† j−1 2{L† jLi, ρP}+O( ˙q3), Li∝Ei,(22) in which only Eiseeds the dissipator; purely central motions (Ei≡0) contribute phases but no collapse [91–95]. The instantaneous collapse (dephasing) rate within the pointer basis is quadratic in the metric speed γ(t) = η˙qi˙qjgijq(t), gij (q) = Tr ∂iP ∂jP,(23) where η is set by relaxation in Q (e.g., Purcell rate in cQED or optical leakage in Λ–systems). Equations (21) – (23) constitute a first–principles derivation linking non–central projector curvature to a CP collapse channel; central curvature (Berry/Wilczek–Zee) alone yields phases without dissipation [96–99]. (C2) Intuitive “shadow” picture and measurable implications. We provide a kinematic interpretation in which the relevant rug P is dragged through Hilbert space while the token (state) develops a shadow off the rug, Q˙ P P|ψi . For rank–1 dressed sectors (Bloch sphere), the complex shadow amplitude is h−n|˙ +ni=1 2˙ θ+i 2˙ φsin θ, h−n|˙ +ni2=1 4˙ θ2+ sin2θ˙ φ2,(24) so that the shadow’s magnitude is exactly the metric speed on the Bloch sphere (Fubini–Study line element) [ 100 , 101 ]. This amplitude is measurable: in cQED it appears as a controllable change in output photon flux under heterodyne/homodyne monitoring; in general it manifests in the statistics of quantum trajectories (e.g., click rates), thereby encoding the direction and speed of subspace motion [ 102 – 106 ]. The picture clarifies the role of the environment: a lossless Q allows, in principle, reversibility of leakage; alossy Qturns the same shadow into irreversible collapse at a rate fixed by η. (C3) Application to nuclear fission: fragment–identity collapse with path dependence. We construct P ( q )as the projector onto fragment subspaces defined by a moving dividing surface (the evolving neck) in nuclear shape space q (elongation, neck radius, asymmetry). The associated metric gij ( q )spikes at the neck, predicting the locus of collapse (where fragment identities become definite) and a threshold law for its timing via R˙qi˙qjgij dt≈ Θ. Because g and the nonabelian curvature F vary across shape space, two protocols with identical endpoints (e.g., elongation → neck vs. asymmetry → neck) acquire different collapse exponents Γ, hence different effective weights Wµ∝exp{−R Γ µdt} on shell–favored partitions, leading to path–dependent widths/centroids in the fragment–mass distribution [ 107 – 111 ]. This goes beyond standard liquid–drop + shell–correction models (which capture asymmetric peaks at A≈95/137 but lack protocol dependence): HCCB pinpoints the basis (fragment numbers), the locus (neck), the timing (threshold), and predicts order–of–operations effects as a falsifiable signature. (C4) Outline of quantum computing applications. We define a collapse functional on control trajectories, JHCCB[q(·)] = ZT 0 η˙qi˙qjgijq(t)dt, (25) and propose compiler/hardware procedures that (a) minimize JHCCB during gate operations to suppress collapse (noise–resilient gates), (b) maximize it in short windows near readout to accelerate basis–aligned measurement, and (c) diagnose geometric collapse by comparing Γfor two protocols with identical endpoints but reversed order (area–law difference) [ 106 , 112 – 117 ]. These prescriptions are platform– agnostic (superconducting qubits, trapped ions, NV centers), and their distinctive scaling with speed (Γ ∝T−1 for fixed rectangle), engineered loss (Γ ∝η ), and enclosed area (order effect) provides a clear experimental path to validation. Summary. Together, (C1)–(C4) elevate collapse from a postulate or noise fit to a computable, structural inevitability of non–central projector geometry. They supply new predictions (basis–locus–timing & protocol dependence) and controls (path, order, speed), with concrete consequences for fission physics and quantum processors. 8 II. THEORETICAL BACKGROUND A. Collapse in standard quantum mechanics Roadmap. We review the standard account of quantum measurement and decoherence with an eye toward what it does and does not predict. We begin with the measurement postulate (projective and generalized/POVM measurements), then derive reduced dynamics and decoherence from explicit system– environment couplings (including exactly solvable models and the Markovian GKSL limit), and finally summarize the limitations: the basis,locus, and timing of collapse are not predicted intrinsically by these frameworks. 2.1.1 Projective measurements (PVM) and Lüders update Spectral calculus and outcome probabilities. Let A be a self–adjoint observable on a separable Hilbert space H with spectral measure EA ( · )(projection–valued measure, PVM). For discrete spectra A = Paa Π a with pairwise orthogonal projectors Πa, the Born rule gives p(a) = Tr Πaρ,X a Πa= 1.(26) Lüders post–measurement state. Conditioned on outcome a, the state update is (Lüders rule [118]) ρ7−→ ρ|a=ΠaρΠa Tr(Πaρ).(27) Ignoring (coarse–graining over) outcomes produces the non–selective update MPVM(ρ) = X a ΠaρΠa.(28) Complete positivity and idempotence (proof). Equation (28) is a completely positive, trace–preserving (CPTP) map with Kraus operators {Ma = Π a} , since PaM† aMa = Pa Π a = 1 and MPVM ( ρ ) = PaMaρM† a is of Kraus form [ 119 , 120 ]. Moreover M2 PVM = MPVM (idempotent), i.e. it is a conditional expectation onto the algebra block–diagonal in the { Π a} basis. Thus projective measurement selects a basis by fiat: the pointer projectors are {Πa}. 2.1.2 Generalized measurements (POVMs), instruments, and dilations POVMs and instruments. Ageneralized measurement with outcomes m is specified by a POVM {Em} (positive operators summing to identity) and an instrument {Im} of CP maps such that p ( m ) = Tr ( Im ( ρ )) and the non–selective channel is CPTP [121–123]. Every instrument admits a Kraus form Im(ρ) = X k Mmk ρ M† mk,X m,k M† mkMmk = 1, Em=X k M† mkMmk.(29) Naimark–Stinespring dilation (construction). Any POVM (and instrument) arises from a unitary interaction with an ancilla and a projective measurement on the ancilla (Naimark) and, conversely, any CPTP map has a dilation (Stinespring) [ 123 – 125 ]: let K be an ancilla Hilbert space, U a unitary on H ⊗ K , and {|mi} an orthonormal basis on K with initial ancilla state | 0 i . Define Kraus operators Mm := hm|U| 0 i ; then (29) follows and Em = M† mMm .Proof sketch: Stinespring asserts that any CPTP map Φ( ρ ) = PjVjρV † j can be realized as Φ( ρ ) = TrK ( V ρV † )for an isometry V : H → H⊗K . Choosing an orthonormal basis on K yields V = PjVj⊗ |ji , from which the above construction is immediate [119, 120]. Consequence. Within the measurement–as–POVM paradigm, any basis can be realized by choosing the coupling and the ancilla measurement appropriately. Thus, while the formalism is universal, it does not predict which basis is preferred; it only constrains dynamical consistency (CP, TP). 9 2.1.3 Decoherence from explicit system–environment models Generic structure. Let H = HS + HE + HI with HI = PαAα⊗Bα . Starting from a product state ρS (0) ⊗ρE , the reduced dynamics is ρS ( t ) = TrEU ( t ) ρS (0) ⊗ρEU† ( t )  . In a nondemolition/pure– dephasing model ([ HS, Aα ]=0), one obtains exact decoherence functions; in more general models, Markovian master equations arise under Born–Markov–secular approximations [127–129]. Exactly solvable pure dephasing (spin–boson). Take HS = ~ω0 2σz , HE = Pk~ωkb† kbk , and HI = σzPk~ ( gkb† k + gkbk ), with ρE thermal. The unitary evolution displaces each bath mode conditionally on the qubit state, yielding for the off–diagonal element (interaction picture): ρ01(t) = ρ01(0) e−iω0te−Φ(t),Φ(t)=2Z∞ 0 dω J(ω) cothβ~ω 2 1−cos ωt ω2,(30) where J ( ω ) = Pk|gk|2δ ( ω−ωk )is the spectral density [ 130 ]. Decoherence is complete if the bath generates orthogonal environmental states conditioned on σz = ± 1. The pointer basis is the eigenbasis of the coupling operator ( σz here) by assumption; change the coupling to σx and the pointer basis changes accordingly. Quantum Brownian motion (QBM) and spatial decoherence. For a particle of mass m in a high– temperature Ohmic bath, the Caldeira–Leggett master equation (in position representation) reads [131, 132] ∂ ∂tρ(x, x0, t) = −i ~hx|[HS, ρ]|x0i− iγ ~(x−x0)∂ ∂x −∂ ∂x0ρ−2mγkBT ~2(x−x0)2ρ, (31) with damping γ . The last term suppresses coherences over spatial separation ∆ x = x−x0 at a rate Γ ϕ∼2mγkBT ~2 (∆ x ) 2 , recovering the intuitive scaling that macroscopic superpositions decohere rapidly. Again, the basis (here, approximate position eigenstates) is induced by the form of the coupling ( x to the bath); the formalism does not derive it from control geometry. 2.1.4 From Redfield to GKSL: Markovian master equations Born–Markov–secular derivation (outline). In the interaction picture, ˙ρI SE ( t ) = −i ~ [ HI I ( t ) , ρI SE ( t )]. Integrating and iterating to second order, tracing over E , and assuming (i) Born ρI SE ( t ) ≈ρI S ( t ) ⊗ρE , (ii) Markov ρI S ( t−τ ) ≈ρI S ( t )and fast decaying bath correlations, and (iii) secular (rotating–wave) approximation in the eigenoperators Aα(ω)of HS, one obtains the GKSL generator ˙ρS(t) = −i ~[HS+HLS, ρS] + X ωX α,β γαβ(ω)Aβ(ω)ρSA† α(ω)−1 2{A† α(ω)Aβ(ω), ρS},(32) where γαβ ( ω )are the Fourier transforms of bath correlation functions and HLS is the Lamb shift [ 127 , 128 , 133 ]. The pointer basis is typically the joint eigenbasis of the Lindblad operators in the dissipator; but these depend on the chosen system–bath coupling HI and on the secular decomposition. Thus, again, the basis is posited, not predicted intrinsically. Unravelings and measurement schemes. The same master equation (32) admits many stochastic unravelings (quantum trajectories). Direct photo–detection (quantum jumps) and homodyne/heterodyne detection (diffusive trajectories) correspond to inequivalent conditional evolutions, even though the ensemble average E [ ρc ( t )] reproduces (32) [ 129 , 134 , 135 ]. Hence, the conditioned “collapse” basis depends on the measurement scheme, not only on the system. This illustrates that the standard formalism, while consistent, does not single out a unique preferred basis. 2.1.5 Zeno limits and continuous monitoring Zeno scaling. Frequent projective measurements of a projection P (interval τ ) give survival probability Psurv ( t ) ≈ 1 −(∆H)2 ~2τ2t/τ −−−→ τ→0 1, freezing evolution (quantum Zeno effect) [ 136 , 137 ]. Collapse here is enforced by measurement frequency, not dynamically predicted. Continuous weak measurement yields stochastic master equations whose diffusion or jump character is set by the detector [ 129 , 138 ]. Again, the locus/timing of “collapse” are measurement–protocol properties. 16 Lemma II.1 (Metric identity).For any trajectory q(t), Tr(˙ P)2= 2 Tr E†E= ˙qi˙qjgij(q), gij (q) := Tr(∂iP ∂jP).(59) In particular, for rank-1projectors P=|+nih+n|on the Bloch sphere with angles (θ, φ), h−n|˙ +ni=1 2˙ θ+i 2˙ φsin θ, h−n|˙ +ni2=1 4˙ θ2+ sin2θ˙ φ2=1 4k˙ nk2,(60) and gθθ =1 2,gφφ =1 2sin2θ,gθφ = 0. Proof. Using P˙ PP = 0 and Q˙ PQ = 0, the tangent ˙ Pis purely off–diagonal: ˙ P=E+E†. Then Tr(˙ P)2= Tr(E†E) + Tr(EE†) = 2 Tr(E†E), which equals ˙qi˙qjTr ( ∂iP ∂jP )by bilinearity and the definition of Ei . For rank-1, insert | + ni = (cos θ 2, eiφ sin θ 2)>and |−ni= (−sin θ 2, eiφ cos θ 2)>, differentiate, and take overlaps.  Interpretation. Lemma II.1 shows that the size of the shadow is fixed by the metric speed k˙qk2 g = ˙qi˙qjgij . This identity is kinematic and independent of any bath; it holds whenever P moves in the Grassmannian. 2.4.2 Reduced dynamics with a fast complement Assumptions. We work under the following hypotheses: (H1) (Spectral regularity) P ( q )is twice continuously differentiable and arises (e.g.) from a gapped spectral cluster of an instantaneous Hamiltonian H(q). (H2) (Separation of timescales) The Q –sector relaxes rapidly to a unique stationary state under a Liouvillian LQwith gap κ > 0(e.g., radiative loss, engineered dissipation). (H3) (Slow control) The control velocity scale satisfies k˙qk  κ/ Λ, where Λbounds operator norms of ∂iPand the P/Qblock couplings (ensuring a regular adiabatic/TCL expansion). Time–local elimination (sketch). Partition the total Liouvillian L ( t )in P/Q blocks and apply the time–convolutionless (TCL) projection–operator method to second order in ˙q while resumming Q –sector fast relaxation. Solving the Q –equation to leading order in 1 /κ and feeding back into the P –equation yields a CP, time–local master equation for ρP=PρP of the form ˙ρP=−i ~[Heff (t), ρP] + X i,j ˙qi˙qjLiρPL† j−1 2{L† jLi, ρP}+O˙q3 κ2,(61) with jump operators proportional to the non–central derivatives: Li=√η Ei, Ei:= Q ∂iP P. (62) Here η∼κ−1 Φis a dimensionful factor set by the Q –sector propagator and its coupling to dissipation, and Heff contains the projected Hamiltonian together with central connection (parallel–transport) terms. The dissipator in (61) is seeded exclusively by Ei ; purely central motions [ Ki, P ]modify Heff but do not generate dissipation at this order. Rigorous adiabatic bounds for such eliminations (closed and open systems) can be established under (H1)–(H3) [171–175]. 2.4.3 Collapse rates and bounds on coherences Instantaneous rate. Let {Pµ ( q ) } be a smooth block–resolution of P ( q )(candidate pointer sectors) and define ρµν =PµρPPνfor µ6=ν. Projecting (61) onto the µν block gives d dtkρµν k2 2=−2γµν(t)kρµνk2 2+skew–Hermitian terms, γµν (t) = ˙qi˙qjΞ(ij) µν (q(t)) ≥0,(63) where Ξ(ij) µν =ηhµ|∂iP ∂jP|µi+hν|∂iP ∂jP|νi−2 Re hµ|∂iP|νihν|∂jP|µi.(64) 17 Summing over µ6=νand using Pµ6=νΞ(ij) µν ≤η gij yields the scalar bound γ(t)≤η˙qi˙qjgijq(t)(65) with equality for rank–1 or for block choices that diagonalize the quadratic form. Thus, the instantaneous collapse rate is controlled by the metric speed of P. Integrated bound and Grönwall inequality. Define the integrated collapse exponent Γ[q(·)] := Zt t0 γ(τ)dτ. (66) Equation (63) implies kρµν(t)k2≤e−Γµν [t0 →t]kρµν(t0)k2,Γµν[t0→t] = Zt t0 γµν(τ)dτ, (67) and therefore (using (65)) kρµν(t)k2≤exp−Zt t0 η˙qi˙qjgij(q(τ)) dτkρµν(t0)k2.(68) Inequality (68) is a direct application of Grönwall–Bellman bounds for linear differential inequalities [ 176 ]. Geometric threshold law. Given a target visibility tolerance 0<ε<1, define Θ(ε) := log1 ε.(69) If along the realized trajectory q(t)an interval [t0, t?]satisfies Zt? t0 η˙qi˙qjgij(q(t)) dt ≥Θ(ε),(70) then (68) ensures kρµν ( t? ) k2≤εkρµν ( t0 ) k2 for all µ6 = ν . Equivalently: collapse occurs when the integrated metric speed exceeds a threshold, establishing (57). 2.4.4 Path dependence and order effects Area law for protocol pairs. Consider two protocols with identical endpoints and total time T that traverse the edges of a small rectangle Rin Min opposite orders (1→2vs. 2→1). A Taylor expansion of gij and Stokes’ theorem yield Γ1→2−Γ2→1≈η TZZR∂q1g22 −∂q2g11dq1dq2=: η TZZRKdS, (71) an order–dependence proportional to an effective curvature K built from the metric (covariant expressions relate K to the projector curvature for composite subspaces). Hence, even at fixed endpoints and duration, the integrated collapse exponent depends on the path, furnishing a falsifiable signature. 2.4.5 Special cases and sharpness Rank–1 (two–level) equality. For rank–1 P, (65) is an equality: γ(t) = η˙qi˙qjgij(q(t)) = ηh−n|˙ +ni2, and (68) integrates exactly to |ρ+−(t)|=e−Γ[t0→t]|ρ+−(t0)|,Γ = Zγ dt. This provides a clean platform for experimental tests in driven qubits (e.g., superconducting circuits). 18 Reversible leakage vs. irreversible collapse. If the Q –sector is lossless and remains coherently trackable, then E6 = 0 still produces reversible leakage; the CP dissipator of (61) emerges only when Q is fast and relaxing (assumption H2). Thus, geometry (through E and g ) sets the propensity for collapse; the environment sets its irreversibility scale η. Gauge invariance. The rate γ and the threshold integral are gauge–invariant: they depend only on P through gij = Tr(∂iP ∂jP), not on a particular choice of frame inside Ran P. 2.4.6 Statement of the collapse theorem Theorem II.2 (Collapse from non–central projector motion) . Let q7→ P ( q )be C2 and satisfy (H1)–(H3). Along any differentiable control path q ( t ), the reduced state ρP = PρP evolves according to the CP master equation (61) with jump operators Li = √η Ei and instantaneous dephasing rate bounded by (65) . Consequently, for any pair of blocks µ6=νin a smooth resolution of P, kρµν(t)k2≤exp−Zt t0 η˙qi˙qjgij(q(τ)) dτkρµν(t0)k2.(72) In particular, if the threshold condition (70) holds on [ t0, t? ], then kρµν ( t? ) k ≤ εkρµν ( t0 ) k for all µ6 = ν . Moreover, for two protocols with the same endpoints and duration enclosing a small area R , the difference of integrated collapse exponents obeys the area law (71). Proof sketch. The CP form (61) with Li∝Ei follows from TCL/Feshbach elimination under (H2)– (H3), using that Q˙ PP = ˙qiEi is the only kinematic source coupling P to Q at order ˙q ; second–order contributions produce the dissipator with coefficient η∼κ−1 Φ. Projection onto off–diagonals yields (63) ; summing and bounding by gij gives (65) . Grönwall–Bellman then yields the exponential bound and the threshold law (70) . For the area law, Taylor–expand gij over R and integrate along the two orderings to obtain (71) . Rigorous adiabatic error bounds for the underlying elimination appear in [ 171 , 174 ], while open–system adiabatic theorems justify the CP reduction in [172, 173].  Consequences. Theorem II.2 elevates collapse from an environmental postulate to a geometric threshold phenomenon governed by the metric of the projector bundle. It determines: (i) the basis stabilized by collapse (the block structure of P that diagonalizes the quadratic form), (ii) the locus (hot–spots where g spikes), (iii) the timing (threshold integral), and (iv) the protocol dependence (area law). III. THE SHADOW PICTURE A. Rug analogy Idea. The rug analogy provides an intuitive and calculationally exact picture of how a moving relevant subspace causes coherence to “spill” into its orthogonal complement. The rug is the instantaneous relevant subspace Ran P ( q ), represented by an orthogonal projector P ( q ) = P ( q ) † = P ( q ) 2 on Hilbert space H ; the token is the physical state |ψi ∈ H ; the shadow outside the rug is its component Q ( q ) |ψi with Q ( q ) := 1−P ( q ). As the rug is dragged over the floor of Hilbert space by changing controls q ( t ), the token—even if it does not move dynamically—casts a time–dependent shadow off the rug. The size of the shadow is determined kinematically by the quantum metric on the Grassmannian of projectors, while its path dependence (how shadow accumulates under different orders of moves) is governed by the curvature of the projector bundle. 3.1.1 Kinematics of a moving rug Setup and notation. Let q7→ P ( q )be C2 on a smooth control manifold M , and fix a differentiable control path t7→ q(t). Define ˙ P(t) = ∂iPq(t)˙qi(t), Ei(q) := Q(q)∂iP(q)P(q), E(t) := Q˙ P P = ˙qiEi.(73) By differentiating P2 = P , one has P˙ PP = Q˙ PQ = 0, so ˙ P is purely off–diagonal and E is its Q←P block. 19 Shadow with a stationary token (first principles). Imagine the token |ψi is fixed in H while only the rug (the choice of relevant subspace) moves. The shadow relative to the moving rug is |shadow(t)i:= Qq(t)|ψi.(74) For an infinitesimal step δt, Qq(t+δt)|ψi=Q(q)−δt ˙ P(q) + O(δt2)|ψi =−δt Q ˙ P P |ψi+O(δt2) =−δt E |ψi+O(δt2),(75) where we used Q ( q ) |ψi = 0 if initially |ψi ∈ Ran P ( q )and Q˙ P Q = 0. Thus the shadow amplitude created in time δt is exactly E|ψito first order. Lemma III.1 (Quadratic law for the shadow norm) . If |ψi ∈ Ran Pq ( t )  at time t , then the shadow norm after a small step is kQ(q(t+δt)) |ψik2=δt2hψ|E†E|ψi+O(δt3).(76) Proof. Equation (75) gives the leading term −δt E|ψi ; taking the squared norm and expanding to second order yields the claim.  Metric control of shadow size. By the projector–geometry identity Tr ( ˙ P ) 2 = 2 Tr ( E†E ), the metric speed k˙qk2 g:= ˙qi˙qjgij(q), gij (q) = Tr ∂iP ∂jP,(77) controls the typical squared shadow per unit time: for an ensemble uniformly distributed in Ran P , E[kE|ψik2] = 1 rTr(E†E) = 1 2rk˙qk2 gfor rank r= rank P. Rank-1 explicit formula (Bloch sphere). For P=|+nih+n|with spherical angles (θ, φ)one has h−n|˙ +ni=1 2˙ θ+i 2˙ φsin θ, h−n|˙ +ni2=1 4˙ θ2+ sin2θ˙ φ2.(78) The real part measures meridional speed ( ˙ θ ) and the imaginary part measures zonal speed ( ˙ φsin θ ). The total magnitude equals the Fubini–Study line element (half the metric speed) [177]. 3.1.2 Moving rug with a dynamically evolving token Coupled kinematics. Let the token |ψ ( t ) i evolve under the Schrödinger equation i~˙ ψ = H ( q ( t )) |ψi . Define components |ψPi=P|ψiand |ψQi=Q|ψi. Differentiating and using P˙ PP =Q˙ PQ = 0, ˙ |ψQi=−(˙ P)P|ψi− i ~QH|ψi=−E|ψPi− i ~QH|ψPi+|ψQi,(79) ˙ |ψPi= ( ˙ P)Q|ψi− i ~PH|ψi=E†|ψQi− i ~PH|ψPi+|ψQi.(80) Equations (79) – (80) show that only the non–central shear E = Q˙ P P injects shadow proportional to the control velocity. The Hamiltonian PHQ also mixes P and Q , but its effect is independent of the rug’s velocity; isolating velocity–induced shadow singles out E. Fast complement and effective dissipation. If the Q –sector is rapidly relaxing (Liouvillian gap κ ) while P is slow, adiabatic elimination yields a completely positive (CP) reduced master equation on ρP = PρP with jump operators proportional to Ei , as derived in Sec. II D. The shadow flux into Q per unit time is then proportional to Tr(E†E ρP), and the collapse rate inside Pbecomes γ(t) = η˙qi˙qjgij(q(t)),(81) with η set by Q ’s relaxation scale. Thus, when the token also moves dynamically, the rug’s geometry still controls the velocity–induced shadow. 20 3.1.3 Path dependence: curvature as an “order memory” of the shadow Order of moves and curvature. Consider two small displacements δq1, δq2 from q to q + δq1ˆe1 + δq2ˆe2 . The final projector P ( q + δq1, δq2 )is path–independent, but the history of shadow production depends on the order in which the rug was dragged. For a stationary token (to isolate geometry), integrate the instantaneous shadow norm along the two orderings. Expanding gij to first order and integrating yields the area law Γ1→2−Γ2→1≈η TZZR∂q1g22 −∂q2g11dq1dq2=η TZZRKdS, (82) where Γis the integrated collapse exponent and R is the rectangle traced by the two segments in time T . The scalar K is a curvature–like density built from the metric g , and in a covariant gauge it can be expressed through the projector curvature two–form F = P [ ∂1P, ∂2P ] P (nonabelian Stokes–type relations) [ 178 , 179 ]. Equation (82) states: the shadow remembers the order of moves—a geometric, path–dependent signature. Rank-1 demonstration on the Bloch sphere. Let q = ( θ, φ )and choose two protocols with identical endpoints (θ0, φ0)→(θf, φf)and total time T: A. θ-ramp at fixed φ=φ0for T/2, then φ-ramp at fixed θ=θffor T/2. B. φ-ramp at fixed θ=θ0for T/2, then θ-ramp at fixed φ=φffor T/2. From (78) , the instantaneous shadow magnitude is 1 2p˙ θ2+ sin2θ˙ φ2 . With constant segment speeds, the integrals give ΓA=η"2 TZθf θ0 1 4dθ +2 TZφf φ0 1 4sin2θfdφ#=η 2T∆θ+ sin2θf∆φ,(83) ΓB=η 2Tsin2θ0∆φ+ ∆θ,(84) hence ΓA−ΓB=η 2T(sin2θf−sin2θ0) ∆φ≈η T(sin θcos θ) ∆θ∆φ, (85) which matches the area law (82) with K = ∂θgφφ −∂φgθθ = 1 2sin θcos θ . Thus, even for a qubit, shadow accumulation is order–dependent. 3.1.4 Observables of the shadow Shadow as a measurable flux. When the complement Q is lossy/fast, adiabatic elimination gives a CP reduced equation with jump operators Li = √η Ei . The expected jump rate (number of quanta emitted per unit time into the environment) is R(t) = X i ˙qi˙qjTr E† iEjρP(t)≤η˙qi˙qjgij(q(t)),(86) so monitoring the environment (photon counts, phonon emission, etc.) gives a direct readout of the metric speed and, by protocol comparison, of the order–dependence (82) . For mixed states, an analogous statement holds with the Uhlmann connection replacing Berry’s phase and with the Bures metric providing the relevant distance—again tying shadow to differential geometry [180, 181]. Gauge invariance of the shadow record. Although frames inside Ran P can be changed by U ( r ) rotations (gauge), the shadow operator E = Q˙ P P and its norm Tr ( E†E )are gauge–invariant, depending only on the projector P and its derivative. Hence the shadow is a physical record of how the rug was dragged, not an artifact of coordinates. 21 3.1.5 Summary In the rug analogy, the projector subspace is the rug, the state is the token, and the shadow is the component in the orthogonal complement. The instantaneous shadow amplitude is exactly E = Q˙ P P , and its average squared norm is proportional to the metric speed k˙qk2 g . Over protocols with identical endpoints but different order, integrated shadow differs by an area law set by curvature–like derivatives of g (and covariantly by the projector curvature F ). When the complement is lossy, this shadow becomes an irreversible record in the environment, with a rate proportional to ηk˙qk2 g . Thus the rug analogy is not merely pictorial: it is a faithful translation of projector geometry into measurable fluxes. B. Shadow in standard quantum mechanics Roadmap. In the standard formalism, the “shadow” of a quantum state does not arise as a geometric quantity attached to a moving relevant subspace, but as the environmental record created by entanglement between the system and its surroundings. This shadow is typically inaccessible: when one traces out the environment, one observes only statistical loss of phase coherence (decoherence) in the reduced density operator. In this subsection we (i) derive the structure of the environmental shadow as overlaps of environment conditionals, (ii) connect it to the influence functional and to noise spectral densities via cumulants/filter functions, (iii) discuss trajectory/unraveling views and collisional models, and (iv) emphasize why, in this account, the shadow is modeled statistically and does not intrinsically encode basis/locus/timing the way projector geometry does. 3.2.1 Environment entanglement as the shadow Measurement-like coupling and conditional environment states. Let HS be the system Hilbert space with an orthonormal basis {|ai} , and let HE be the environment. Consider a measurement-like interaction in which HI=X a|aiha| ⊗ Ba, UI(t) = e−i ~HIt=X a|aiha| ⊗ e−i ~Bat.(87) With an initial product state ρ(0) = ρS(0) ⊗ρE, the joint state at time tis ρSE(t) = X a,b ρab(0) |aihb| ⊗ e−i ~BatρEei ~Bbt,(88) and the reduced system state is ρS(t) = X a,b ρab(0) Dab(t)|aihb|, Dab(t) := TrEe−i ~BatρEei ~Bbt.(89) If ρE = |0ih0| is pure, then Dab ( t ) = hb ( t ) |a ( t ) i with |a ( t ) i := e−i ~Bat|0i . Thus the off-diagonal coherence ρab ( t )equals the overlap of environment shadow states conditioned on a and b . When |a ( t ) i and |b ( t ) i become orthogonal, Dab ( t ) → 0and the corresponding coherence vanishes. This is einselection: pointer states are eigenstates of the coupling operator [186, 187]. General Hamiltonians and Loschmidt echo. For a general H = HS + HE + HI , one can write (for pure |Ψ(0)i=Paca|ai⊗|0i) |Ψ(t)i=X a ca|ai⊗|a(t)i,|a(t)i=Te−i ~Rt 0ha|H|aidτ |0i,(90) whence the reduced coherence is ρab ( t ) = cac∗ bhb ( t ) |a ( t ) i . The factor Mab ( t ) := hb ( t ) |a ( t ) i is a Loschmidt echo (fidelity amplitude) between two environmental evolutions; its decay underlies decoherence [ 188 ]. In this view, the shadow is precisely the set of environment conditional states {|a ( t ) i} , which is typically unobserved, hence traced over. 22 3.2.2 Influence functional and cumulant expansions Feynman–Vernon influence functional. In the path-integral representation for a system coordinate x ( t )linearly coupled to a bosonic bath, integrating out the bath yields an influence functional F [ x, x0 ] such that ρS(xf, x0 f;t) = ZDxDx0ei ~(S[x]−S[x0]) F[x, x0]ρS(x0, x0 0; 0).(91) For a Gaussian bath, F[x, x0] = exp−1 ~Zt 0Zt 0 dτ dτ0ξ(τ)ν(τ−τ0)ξ(τ0)−i ξ(τ)µ(τ−τ0)χ(τ0),(92) with sum/difference coordinates χ = 1 2 ( x + x0 ), ξ = x−x0 , and bath noise/dissipation kernels ν and µ fixed by the spectral density and temperature [ 189 , 190 ]. The decoherence functional is the real exponential in (92) , which suppresses contributions with large separations ξ ; it is entirely a property of the bath statistics. Again, the shadow is represented implicitly by F. Cumulant and Gaussian approximation. For stationary, Gaussian fluctuations of a (possibly classical) stochastic process β(t)that couples linearly to an observable Avia HI=β(t)A, one obtains hTe−i ~Rt 0β(τ)AI(τ)dτ i= exp−1 2~2Zt 0Zt 0 dτ dτ0C(τ−τ0)h{AI(τ), AI(τ0)}ic,(93) where C ( τ ) = hβ ( τ ) β (0) i is the classical autocorrelation and h·ic indicates connected correlators. For pure dephasing of a qubit ( A = ~ 2σz ) with toggling function y ( t ) ∈ {± 1 } (due to control), the off-diagonal element obeys ρ01(t) ρ01(0) =Deiφ(t)E= exp−1 2hφ2(t)i, φ(t) = Zt 0 y(τ)β(τ)dτ. (94) Equation (94) is the basic Gaussian dephasing law; it expresses decoherence entirely through the noise statistics [191, 192]. 3.2.3 Spectral densities and filter functions Filter-function formalism. Assuming β ( t )is stationary with one-sided power spectral density S ( ω ) = 2R∞ 0C(τ) cos(ωτ)dτ, the variance in (94) becomes hφ2(t)i=1 πZ∞ 0 dω ω2S(ω)|F(ωt)|2, F (ωt) = ωZt 0 y(τ)eiωτ dτ, (95) so that ρ01(t) ρ01(0) = exp−1 2πZ∞ 0 dω ω2S(ω)|F(ωt)|2.(96) Equation (96) captures the standard doctrine: the shadow (decoherence) is modeled statistically via the bath spectrum and shaped by control through a filter function F . It underlies noise spectroscopy with dynamical decoupling and coherence tailoring [ 193 – 195 ]. In solid-state qubits, specific forms like 1 /f noise (S(ω)∝1/ω) or Ohmic noise yield characteristic decay envelopes [196, 197]. 3.2.4 Trajectories, input–output, and inaccessibility of the shadow Quantum trajectories and unravelings. The same reduced master equation may be unraveled into stochastic trajectories conditioned on continuous measurements of the output field. Quantum-state diffusion (QSD) and quantum-jump formalisms produce diffusive and jump dynamics for the conditioned state |ψc ( t ) i [ 198 ]. In input–output theory, the output operator bout ( t ) = bin ( t ) + √κ c ( t )carries away the environmental record (the shadow) in propagating modes [ 199 , 200 ]. Unless these modes are monitored, one averages over them, and the shadow remains inaccessible, appearing only as ensemble dephasing. Different detection schemes (homodyne vs. direct detection) lead to different conditionals but the same average decoherence; thus, within standard QM, the shadow’s structure is external to the system and specified by the measurement choice. 23 3.2.5 Collisional decoherence: scattering as shadow formation Momentum kicks and spatial decoherence. When a mesoscopic particle scatters off a dilute gas, each collision imprints which-path information on the gas; tracing over gas modes yields spatial decoherence with a rate that scales with the square of the separation and the gas cross section. In the Markovian limit, the master equation in position representation reads ∂ ∂tρ(x, x0;t) = −Λ1−Φ(x−x0)ρ(x, x0;t),(97) where Φis the elastic scattering form factor and Λdepends on gas density and cross sections [ 201 , 202 ]. Again, the shadow—the gas record—is not computed as a geometric object of system control but as a statistical property of the bath. 3.2.6 Why the standard shadow is statistical (and its limits) Summary of structure. Across the models above, the shadow of a quantum state in standard QM is the ensemble of environmental conditionals {|a ( t ) i} (or field output modes) produced by the interaction. After tracing out the environment, the reduced state exhibits dephasing controlled by: (i) overlaps of these environment states (Loschmidt echo), (ii) influence functionals determined by bath correlations, or (iii) noise spectral densities filtered by control protocols. In all cases, the shadow is extrinsic to the system: it lives in the environment and is typically modeled statistically. No intrinsic basis/locus/timing. Because the shadow is defined by the chosen coupling to the environment and by the measurement (or lack thereof) of the environment, the standard account does not intrinsically predict which basis collapses (it depends on HI or detection scheme), where along a driven path collapse occurs (it depends on thresholds or detector bandwidths), or when collapse occurs (it is set by noise spectra and dwell times). Non-Markovian refinements quantify information backflow but retain the same extrinsic status for the shadow [203, 204]. Contrast to the geometric (HCCB) picture. By comparison, the HCCB framework treats the shadow amplitude as a geometric quantity E = Q˙ P P controlled by the motion of the projector; the size is fixed by the metric and the path dependence by curvature. Hence HCCB internalizes the shadow into the system’s control geometry, yielding structural predictions absent from the standard, environment-statistical approach. IV. PATH-DEPENDENT COLLAPSE A. Shadow in HCCB Idea. In the HCCB framework, the shadow is not an implicit, statistical by–product of tracing out an environment. It is an explicit kinematic vector in the orthogonal complement of the relevant subspace, determined solely by how the projector (the “rug”) moves in Hilbert space. Given a smooth map q7→ P ( q ) and a control path q(t), the shadow operator is E(t) := Q(q(t)) ˙ P(q(t)) P(q(t)) = ˙qi(t)Eiq(t), Ei(q) := Q(q)∂iP(q)P(q), Q =1−P. (98) For any normalized state |ψi ∈ Ran Pq(t), the shadow amplitude is the vector |shadow(t)i=E(t)|ψi ∈ Ran Qq(t),(99) and its squared norm exactly quantifies the instantaneous leakage created by the motion of the projector. 3.3.1 Geometry fixes the shadow: metric and curvature Metric control (size). Using P˙ PP = Q˙ PQ = 0, one has ˙ P = E + E† . The quantum (Grassmannian/Fubini–Study) metric reads gij(q) = Tr ∂iP ∂jP= 2 Re Tr E† iEj.(100) 24 Hence the metric speed k˙qk2 g= ˙qi˙qjgij satisfies k˙qk2 g= Tr (˙ P)2= 2 Tr E†E,hshadow|shadowi=hψ|E†E|ψi.(101) Averaged uniformly over |ψi ∈ Ran Pof rank r,Ehshadow|shadowi=1 2rk˙qk2 g. Thus, size is fixed by g. Curvature control (order dependence). For two control directions q1, q2, the projector curvature F12(q) = P[∂1P, ∂2P]P(102) governs the second–order failure of displacements to commute: δP(1→2) −δP(2→1) =1 2[∂1P, ∂2P]δq1δq2+O(δq3). Consequently, the integrated shadow (and any observable built from it) differs for the two orderings by an area law Γ1→2−Γ2→1≈η TZZR∂q1g22 −∂q2g11dq1dq2,(103) which can be written covariantly in terms of F12 for composite subspaces (nonabelian Stokes–type relations). Thus, direction and protocol are encoded by F. 3.3.2 From shadow to collapse: jump rate and counting statistics Reduced dynamics with a fast complement. Let the Q sector relax rapidly (Liouvillian gap κ ) while P is slow. Time–local elimination yields a completely positive (CP) master equation on ρP = PρP with jump operators Li(q) = √η Ei(q), η ∼κ−1×(fast propagator factors),(104) leading to the dissipator Pi,j ˙qi˙qjLiρPL† j−1 2{L† jLi, ρP} . The only seed of dissipation from projector motion is Ei; purely inner motions [Ki, P]induce phases but no jumps. Counting process and shadow measurement. Consider the quantum–jump unraveling for the output channel associated with L=Pi˙qiLi=√η E. The stochastic counting process Nthas intensity λ(t) = Tr L†L ρP(t)=ηTr E†E ρP(t)≤η 2k˙qk2 g,(105) and the expected number of counts in [0, T]is E[NT] = ZT 0 λ(t)dt =ηZT 0 Tr E†E ρP(t)dt. (106) Equation (105) shows: measuring the environment counts measures the shadow. In a cQED device, these are output photons; in fission, they are prompt/scission neutrons: both are proportional to Tr ( E†E ρP ) under appropriate coarse–graining. Likelihood and geometric tomography. For Poissonian detection, the log–likelihood of a path q ( · )given a record {tk}is log L[q(·)|{tk}] = X k log λ(tk)−ZT 0 λ(t)dt. (107) Small deformations q7→ q + δq change λ by δλ = η˙qi˙qjδgij + (state terms) , so repeated runs with known ρP enable tomographic reconstruction of gij locally by modulating ˙q and fitting (107). Protocol pairs (A then B vs. B then A) identify curvature via the area difference (103) in the mean counts. 3.3.3 Rank–1 exemplar: closed formulas Bloch sphere shadow and rate. For a dressed qubit with P=|+nih+n|and control angles (θ, φ), h−n|˙ +ni=1 2˙ θ+i 2˙ φsin θ, λ(t) = ηh−n|˙ +ni2=η 4(˙ θ2+ sin2θ˙ φ2).(108) The expected counts and the integrated collapse exponent coincide up to the state factor: E [ NT ] = Rλ ( t ) dt = η 4R ( ˙ θ2 + sin2θ˙ φ2 ) dt . Two rectangles (same endpoints, reversed order) give different E [ NT ] whenever sin2θvaries across the rectangle—an immediately testable signature. 25 3.3.4 Nuclear analogy: scission neutrons as a shadow readout Projector and shadow at scission. In fission, let P ( q )project onto the fragment subspace determined by a dividing surface (the neck) in shape space q = ( elongation,neck radius,asymmetry, . . . ). As the neck thins, P deforms sharply; the non–central derivative E = Q˙ PP transfers amplitude into Q , which collects continuum/outgoing modes including nascent neutron channels. With a fast sink (optical potential), L=√η E sets the scission emission intensity λscission(t) = ηTr E†E ρP(t),(109) so the time–resolved neutron/gamma counts provide a direct shadow record of how the dividing surface moved. Protocol differences (e.g., biasing asymmetry before necking vs. after) then yield order–dependent total scission counts, mirroring (103). 3.3.5 Structural properties: gauge invariance, additivity, bounds Gauge invariance. Under a smooth change of frame inside Ran P (a U ( r )gauge), P is invariant, hence E = Q˙ PP and Tr ( E†E ρP )are invariant. Thus, shadow observables depend only on geometric data (P, ˙ P). Additivity and convexity. For classical mixtures ρP = Pkwkρ(k) P with wk≥ 0, Pkwk = 1, the intensity is linear: λ ( t ) = Pkwkλ(k) ( t ); the integrated exponent is convex in ρP . Hence shadow–based diagnostics are robust to classical averaging. Speed–limit bound. Because dTr ( ρ2 P ) /dt ≤ − 2 λ ( t ) σcoh for a constant σcoh determined by the minimal coherence weight among off–diagonal blocks, one obtains ZT 0 λ(t)dt ≥1 2σcoh Tr(ρ2 P(0)) −Tr(ρ2 P(T)),(110) linking the shadow budget to purity loss (a dephasing speed–limit inequality). 3.3.6 Practical extraction: protocols and metrology Protocol pairs and curvature maps. On a 2D control plane, implement two rectangles of side lengths (∆ q1, ∆ q2 )with identical endpoints and total time T , traversed in opposite orders. The difference in mean counts, ∆N=E[N(1→2) T]−E[N(2→1) T]≈η π T(∆q1∆q2)K, estimates the local curvature density K , with π = RT 0TrP ρP ( t ) dt/T the average P –weight. Scanning the plane reconstructs a curvature map. Metric spectroscopy. With single–axis ramps at fixed speed ˙qi = v and ˙qj6=i = 0, the count rate is λ ( t ) = ηv2gii ( q ( t )); recording λ along the ramp yields gii directly. Cross–terms gij follow from two–axis ramps with known phase relations of velocities. Metrological gain. Because λ scales as the square of velocity, higher–frequency modulations trade bandwidth for count SNR. The Fisher information for parameters θ entering P ( q ; θ )scales as I ( θ ) ∼ R ( ∂θgij ) ˙qi˙qjdt , showing optimal estimator performance when paths align with directions of largest ∂θg . 3.3.7 Relation to dissipative holonomy and mixed–state geometry Dissipative holonomies. In driven–dissipative settings, adiabatic manipulation of steady subspaces yields holonomies and phases that generalize Berry/Wilczek–Zee to nonunitary dynamics. Here, the same geometric data ( g, F )built from P govern both (i) phase–like effects (central) and (ii) shadow/collapse (non–central), unifying dissipative holonomy with HCCB control. Mixed–state phases. For nonunitary evolutions, kinematic phases for mixed states (Uhlmann, interferometric) depend on purifications and parallel transport. In HCCB, the shadow operator E complements those phases by providing the normed leakage channel that is directly observable in counts, tying mixed–state geometry to measurable fluxes. 32 and governs the asymptotic error exponent for discriminating A from B by any test based on D . Thus, path dependence is operational: if λA6 = λB on a set of nonzero measure, then DKL ( AkB ) > 0and the environment’s record can, in principle, reveal which path was taken. Small–difference (local) expansion. If δλ ( t ) := λA ( t ) −λB ( t )is small compared to ¯ λ ( t ) := 1 2 ( λA + λB ), then DKL(AkB) = 1 2ZT 0 δλ(t)2 ¯ λ(t)dt +O(δλ3),(140) so the distinguishability scales with the L2 –distance between hazards. When λ∝Tr ( E†E ρP )and ρP is held fixed, δλ inherits its sign/structure from the geometric difference in E†E caused by changing the order or shape of the path. 4.3.2 What the record encodes: basis, locus, timing, order Basis (what becomes definite). The pointer labels stabilized by collapse are the blocks of P ( q )that diagonalize the quadratic form generated by Ei = Q∂iP P . The record λ ( t )does not depend on the choice of frame within Ran P (gauge invariance) but does depend on which block structure P chooses. Hence the basis is fixed once the relevant projector is specified by the experiment. Locus (where collapse concentrates). Hot–spots of λ ( t )occur where gij ( q ) ˙qi˙qj is large. Define the propensity density Λ(q; ˙q) = ˙qi˙qjgij(q). Then λ ( t ) ≈η 2 Λ q ( t ); ˙q ( t )  (up to state factors), and the times at which Λpeaks are precisely the loci where the environment’s record is most intense (e.g. the neck in fission, avoided crossings in qubits). Timing (when collapse occurs). Let Θ( ε ) = log (1 /ε )be the target visibility threshold. The time of collapse t?is defined implicitly by the geometric threshold Zt? 0 η˙qi˙qjgij(q(t)) dt = Θ(ε). If the path qis varied by δq with fixed endpoints, the first–order shift obeys δt?=−δΓ[0→t?] γ(t?)=−ηRt? 0(2 gij ˙qiδ˙qj+∂kgij ˙qi˙qjδqk)dt η˙qi˙qjgij(q(t?)) ,(141) which shows explicitly how reordering segments (changing δq, δ ˙qlocally) advances or delays collapse. Order (protocol memory). For two–segment protocols enclosing a small rectangle R with equal edge times, the order effect in the collapse exponent is ΓA−ΓB=η TZZR∂Ωg∆∆ −∂∆gΩΩdΩd∆ + O(k∆k3), so the environment’s record carries a signed–area imprint proportional to a curvature–like density built from g(covariantly expressible via the projector curvature F=P[∂P, ∂P]P). 4.3.3 Directionality of motion: beyond magnitudes Complex shadow amplitude. For rank–1 P = | + nih + n| , the shadow amplitude (the Q←P component of ˙ P) is h−n|˙ +ni=1 2˙ θ+i 2˙ φsin θ. Its magnitude controls the jump rate ( ∝ | · |2 ), but its phase encodes the direction of motion on the Bloch sphere (meridional vs. zonal). Diffusive unravelings (homodyne/heterodyne) measure environment quadratures proportional to hL + L†i or hi ( L†−L ) i ; since L∝E = Q˙ P P , the continuous current carries partial information about the complex shadow and thus the direction of motion, not just its speed. Hence, with appropriate monitoring, the shadow can be used to reconstruct the trajectory (up to gauge) from the output record. 33 4.3.4 Reversible leakage vs. irreversible collapse Two regimes. If the complement Q is lossless and coherently tracked, E6 = 0 causes reversible excursions: the shadow exists but can, in principle, be erased by reversing the path (Loschmidt echo within the enlarged P⊕Q space). If Q is fast and relaxing, the same shadow becomes an irreversible Markov jump with rate (137) ; time–reversal cannot reassemble coherence lost to the environment. In both cases, Γstill integrates the metric speed, but only in the latter does it translate into a permanent, countable record. 4.3.5 Three archetypes (a) Driven qubit (rank–1). With (θ, φ)controls, λ(t) = η 4(˙ θ2+ sin2θ˙ φ2)and Γ[q] = η 4ZT 0 (˙ θ2+ sin2θ˙ φ2)dt. Two gates with the same endpoints and duration but different order (e.g. Ry then Rz vs. Rz then Ry ) yield different Γiff ∂θgφφ 6 = 0 along the enclosed strip; the environment’s record can thus diagnose control geometry. (b) Nuclear fission (composite P ). Let P ( q )project onto fragment subspaces defined by a moving neck in shape space q . Neck formation produces a spike in gij ( q ), localizing the locus of collapse. Biasing asymmetry before vs. after necking reorders the traversal of regions with different g , changing Γ(and the time–resolved scission emission (137) ), hence path–dependent widths/weights in fragment–identity outcomes. (c) Many–body dressed manifolds. For an encoded logical subspace with stabilizer projectors, moving P ( q )by parametric drives in the Hamiltonian induces a nontrivial gij ( q ). Compilers that follow geodesics of g minimize Γ(less collapse during gates), whereas readout sequences that graze g –hot–spots maximize early λ ( t )(faster measurement). Order–reversed calibration loops expose ∂igjj , enabling geometry–aware control. 4.3.6 Information and thermodynamic analogies Action, speed limits, and entropy. Γis a quadratic “kinetic” action on M; (125) implies Γ≥ηLg[q]2 T, a geometric speed–limit: for fixed time budget T , paths with longer g –length incur more collapse. In irreversible regimes, the expected entropy produced in the environment (per unit bandwidth) is monotone in Γ; thus geodesic protocols minimize both shadow counts and entropy export. Identifiability and error exponents. The Chernoff/Hellinger distance for Poisson hazards has the closed form C(A, B) = −logZpdPAdPB=1 2ZT 0pλA(t)−pλB(t)2dt, so the best achievable Bayesian error decays as exp [ −C ( A, B )] in the asymptotic–data regime. Since √λ∝ kEk and kEk2∝ k˙qk2 g ,order effects that modify k˙qkg along the path become exponentially distinguishable in the environment’s record. 4.3.7 Summary: what “history–sensitive collapse” means • The shadow amplitude E = Q˙ P P is the vector that encodes how the relevant subspace is being moved right now. Its norm (via g ) sets the instantaneous rate; its complex phase (accessible in diffusive readout) encodes direction. • The integrated shadow Γ = Rηk˙qk2 gdt is a geometric action: it depends on how the path was taken (shape and order), not just on where it starts and ends. 34 • The environment’s record (counts or continuous currents) is a structured time series that reflects the protocol: it localizes where collapse concentrated (hot–spots), when the threshold was crossed (timing), and which ordering of moves occurred (area law). • In reversible regimes, the shadow is a kinematic diagnostic; in irreversible regimes, it is a persistent, information–bearing trace that enables hypothesis tests between protocols with the same endpoints and duration. In short: HCCB turns collapse into a history–sensitive phenomenon. The shadow amplitude explicitly captures the trajectory, and the environment—by counting or continuously monitoring emissions—acts as a faithful recorder of the path taken through control space. V. APPLICATION: NUCLEAR FISSION A. Standard description Roadmap. We summarize the prevailing (pre–HCCB) theoretical picture of low–energy nuclear fission, with emphasis on (i) how shell corrections in the potential–energy landscape explain the characteristic asymmetric fragment–mass yields near A≃ 95 and A≃ 137 (post–neutron) for 235 U( nth, f ); (ii) how fragment identity is operationally assigned at scission in both macroscopic–microscopic transport models (Langevin/Brownian shape motion) and microscopic TDGCM (generator–coordinate) approaches; and (iii) why these standard frameworks, while quantitatively successful, do not intrinsically predict the locus, timing, or protocol dependence of collapse into fragment identity. 5.1.1 Macroscopic–microscopic baseline: potential energy surfaces and shell corrections Shape coordinates and energy decomposition. Let q = ( q1, . . . , qd )denote collective shape variables (e.g., elongation, mass asymmetry, neck radius, higher–multipole deformations), with d∈ [3 , 5] in typical implementations. The macroscopic–microscopic (mac–mic) energy is written E(q) = ELD(q) + δEshell(q) + δEpair(q),(142) where ELD is the liquid–drop (surface + Coulomb) contribution, δEshell the Strutinsky shell–correction (single–particle quantization around the mean field), and δEpair the pairing correction. In the Strutinsky scheme, for a given shape q , one computes the single–particle spectrum {εi ( q ) } of a mean–field Hamiltonian and defines δEshell(q) = N X i=1 εi(q)−e EN(q),e EN(q) = Z˜ λ −∞ ε˜g(ε;q)dε, (143) with ˜g a smooth level density obtained by Gaussian (or polynomial) smoothing and ˜ λ the smoothed Fermi energy. Pairing is added via BCS or HFB corrections. The resulting potential–energy surface (PES) V(q) := E(q)exhibits valleys and ridges associated with shell closures. Asymmetric valley and 95 / 137 yields. A robust feature of actinide PESs is an asymmetric valley toward scission in which the heavier nascent fragment is stabilized near magic (or quasi–magic) numbers (e.g., the Z = 50, N = 82 region around 132Sn ). Transport along this valley leads to pre–neutron primary masses peaked near ( AL, AH ) ≈ (94 , 140); convolution with prompt–neutron emission shifts the peaks to post–neutron values near ( Apost L, Apost H ) ≈ (95 , 137), consistent with observed mass yields for thermal–neutron–induced fission of 235U. 5.1.2 Transport to scission: Langevin and Brownian shape motion Collective dynamics with dissipation and noise. The evolution of the shape coordinates q ( t )is modeled by a Langevin system on the PES, Mij(q) ¨qj+γij (q) ˙qj+∂iV(q) = ξi(t),hξi(t)ξj(t0)i= 2 Tcoll γij(q)δ(t−t0),(144) 35 with collective inertia Mij (e.g., cranking), friction γij (wall/window), and a fluctuation–dissipation relation at collective temperature Tcoll . In the overdamped (Smoluchowski) regime, inertial terms are neglected, giving ˙qi=−µij(q)∂jV(q) + ζi(t),hζi(t)ζj(t0)i= 2 Tcoll µij(q)δ(t−t0), µ =γ−1.(145) The corresponding probability density f(q, t)satisfies a Fokker–Planck equation, ∂tf=∂iµij∂jV f +Tcoll ∂jµijf.(146) Scission surface and yields. Ascission hypersurface S ⊂ M is introduced by an empirical neck criterion, e.g., cneck(q) = Zρ(r;q)wneck(r)d3r≤c∗,(147) where wneck localizes density in the waist region and c∗ is a threshold. The instantaneous scission flux is ΦS(t) = ZS Ji(q, t)ni(q)dS, Ji(q, t) := −µij(q)∂jV(q)f(q, t)−Tcoll ∂jµijf,(148) with n the outward normal. A fragment property (e.g., mass asymmetry α or integer mass A ) is mapped from shape space at scission by a deterministic function Ξ( q )defined on S (e.g., by partitioning the density by the neck). The pre–neutron yield for label ξis then Ypre(ξ)∝Z∞ 0 dt ZS Ji(q, t)ni(q)δξ−Ξ(q)dS, (149) i.e., the time–integrated scission flux binned by Ξ(q). Post–neutron folding. Post–neutron yields follow by convolving with the multiplicity distribution P(ν|ξ), Ypost(A) = X ν≥0Zdξ Ypre(ξ)P(ν|ξ)δA−(ξ−ν),(150) where νis the number of prompt neutrons emitted by the primary fragment labelled by ξ.[218] 5.1.3 Microscopic TDGCM (GCM+GOA): collective wave dynamics and scission flux Collective Schrödinger equation. In time–dependent generator coordinate methods with the Gaussian– overlap approximation (TDGCM–GOA), the collective wavefunction Ψ(q, t)obeys i~∂tΨ(q, t) = "−~2 2 1 pg(q)∂i pg(q)Bij(q)∂j+V(q)#Ψ(q, t),(151) where Bij is the collective inertia (inverse mass tensor) and g ( q )the metric determinant arising from the GOA kernel. The probability current reads Ji(q, t) = ~ 2iΨ∗Bij∂jΨ−(∂jΨ∗)Bij Ψ, ∂t|Ψ|2+∂iJi= 0.(152) Scission mapping and yields. A scission hypersurface S is again defined by a neck operator (or density criterion). The pre–neutron yield for label ξ= Ξ(q)is the flux through S: Ypre(ξ)∝Z∞ 0 dt ZS Ji(q, t)ni(q)δξ−Ξ(q)dS, (153) identical in structure to (149) with J now the quantum current (152) . Post–neutron folding proceeds as in (150). Connection to shell structure. The PES V ( q )entering (151) originates from an underlying energy density functional (EDF) and inherits shell valleys; the concentration of flux into the asymmetric valley explains the dominance of heavy fragments with A≈ 137 (post–neutron) and complementary lights near A≈95. 36 5.1.4 Coulomb repulsion and kinetic observables (context) Scission kinematics. In scission–point models, the total kinetic energy (TKE) is approximated by Coulomb repulsion at separation Rsc plus pre–scission kinetic energy, TKE ≈ZLZHe2 Rsc + TKEpre,(154) with Rsc inferred from the scission configuration, and ZL,H the fragment charges. The total excitation energy TXE = Q−TKE partitions between fragments and feeds prompt neutrons; this partition controls P(ν|ξ)in (150) and thus the shift from primary to post–neutron peaks. 5.1.5 What is assumed as “collapse” in standard models Operational identification at scission. Equations (149) and (153) implicitly assume that when the evolving configuration crosses the scission surface S , the system is irreversibly assigned afragment identity (mass/charge) via the mapping Ξ( q )and that subsequent dynamics do not reshuffle these labels.[ 219 ] This identification functions as an effective collapse postulate: the flux crossing S is taken as a proxy for the Born probability of ending in the corresponding fragment channel. 5.1.6 Limitations vis-à-vis basis, locus, timing, and protocol dependence Basis (what becomes definite). The basis of collapse is chosen by construction: the labels are those produced by the scission mapping Ξ( q )(e.g., fragment mass from a neck partition). The theory does not derive from first principles that these labels are preferred over others; it encodes them in the definition of Sand Ξ. Locus (where collapse occurs). The locus of collapse is set by an empirical scission criterion (147) (choice of neck operator and threshold c∗ ). While different criteria correlate with physical features (rapid density necking, vanishing nuclear contact), the location of S is not predicted intrinsically by the transport equations (146) or the TDGCM dynamics (151); it is an external surface drawn in M. Timing (when collapse occurs). In transport and TDGCM, the time at which a given trajectory/wavepacket element crosses S depends on drift, diffusion, and the PES, but there is no geometric threshold law tied to the speed of a projector—indeed, no moving projector is identified. Collapse timing is thus a by–product of when flux reaches S , not an intrinsic prediction based on a metric on a projector bundle. Protocol dependence (order effects). Standard formalisms predict yields from the integrated scission flux (149) , (153) . If two control scenarios lead to the same f ( q, t )(or | Ψ( q, t ) |2 ) near S at comparable times (or the same asymptotic flux), then the yields coincide, independent of the order in which intermediate regions of M were traversed. In particular, there is no built–in analogue of an area law or a curvature–controlled order dependence of collapse for protocols with identical endpoints and duration. Any residual path sensitivity appears only through dynamical differences in drift/diffusion histories and does not arise from ageometric property of how a relevant projector is moved. Synthesis. Mac–mic transport and microscopic TDGCM successfully reproduce the shape of mass yields—most notably the asymmetric peaks tied to shell closures—and many correlated observables (TKE, ν ( A )). However, they assume a scission mapping (hence the basis and locus of collapse), and they do not supply an intrinsic, geometry–based rule for the timing of collapse or for protocol dependence at fixed endpoints/time. These are precisely the structural gaps that the HCCB framework is designed to fill in the subsequent subsections. B. HCCB perspective Idea. In HCCB the relevant projector P ( q )encodes fragment identity by partitioning configuration space into left and right spatial domains separated by an evolving neck (a moving interface). As collective controls q = ( q1, . . . , qd )(elongation, neck radius, mass asymmetry, higher multipoles) are driven along a protocol q ( t ), the partition (rug)moves in Hilbert space. The metric of this moving projector quantifies how fast the partition moves; it localizes on the neck surface and determines the instantaneous collapse 37 rate of fragment labels. Collapse occurs once the integrated metric speed across the neck exceeds a threshold. This furnishes a structural (basis–locus–timing) account absent in the standard description. 5.2.1 Projectors from a moving neck (level–set construction) Level–set partition. Let φ : R3×M→R be a smooth boundary function such that, for each shape q∈M, Σ(q) = {x∈R3:φ(x;q)=0} is a smooth embedded surface (the neck), with unit normal n(x;q) = ∇φ(x;q) k∇φ(x;q)k. Define the left/right domains by Ω L ( q ) = {x : φ ( x ; q ) < 0 } and Ω R ( q ) = {x : φ ( x ; q ) > 0 } . Choose a smooth step s : R→ [0 , 1] with s ( −∞ )=1, s (+ ∞ )=0, and derivative s0 compactly supported; fix a small interface thickness ` > 0. On the one–body Hilbert space H1 = L2 ( R3 ), define the smoothed multiplication projectors (P(`) L(q)ψ)(x) = sφ(x;q) `ψ(x), P(`) R(q) = 1−P(`) L(q).(155) These converge (strongly) as `→ 0 + to the sharp domain projectors χΩL(q) and χΩR(q) . On the many– body Fock space, the fragment–label projector P ( q )is the spectral projector onto states whose one–body density is block–diagonal in Ω L/ Ω R ; for the geometric (quadratic) quantities below, it suffices to work with (155) and a weight w(x;q)that accounts for the occupied density (see below). Derivative of the projector (shape calculus). Differentiating (155) with respect to qigives ∂iP(`) L(q) = 1 `s0φ `(∂iφ),(156) as a multiplication operator on H1 . The complement Q = 1−P and the Q←P block (the shadow operator) are therefore localized in a thin slab around Σ( q ); the only motion that matters is the normal motion of the interface. 5.2.2 Metric localized on the neck: coarea reduction Hilbert–Schmidt metric with state weight. For multiplication operators, the canonical Hilbert–Schmidt pairing reduces to an L2 integral. Introducing a positive state weight w ( x ; q )(e.g. a suitable functional of one–body density and pairing fields) to capture occupancy, we define g(`) ij (q) := ZR3 w(x;q)∂iP(`) L(x;q)∂jP(`) L(x;q)d3x. (157) Insert (156), apply the coarea formula ZR3 f(φ(x;q)) h(x)d3x=Z+∞ −∞ ds Z{φ=s} h(x) k∇φ(x;q)kdS(x), with f(u)=[s0(u/`)]2/`2, and let cs:= R+∞ −∞ [s0(u)]2du > 0. In the thin–interface limit `→0+, g(`) ij (q) = cs `ZΣ(q) (∂iφ)(∂jφ) k∇φkw(x;q)dS(x) + O(1).(158) Absorbing the thickness prefactor κ` := cs/` into the overall scale (it is ultimately folded into the environmental factor ηof the collapse rate), we obtain the neck–localized metric gij(q) = κZΣ(q) (∂iφ)(∂jφ) k∇φkw(x;q)dS(x) = κZΣ(q) ui(x;q)uj(x;q)w(x;q)dS(x),(159) 38 where we introduced the normal velocity fields ui(x;q) := −∂iφ(x;q) k∇φ(x;q)k.(160) Thus, gij is precisely the surface average of the product of normal velocities associated with the control directions qiand qj, weighted by w. Remarks on the weight w .For Slater–determinant states, the quadratic pairing (157) reduces to a one–body integral with w ( x ; q ) ≈ 2 ρ ( x ; q )  1 −ρ ( x ; q )  (spin factor 2), where ρ is the occupation density in the interfacial region. In superfluid HFB, w should incorporate both normal and anomalous densities; the above derivation, however, is agnostic to the detailed choice, as long as w≥0and smooth. 5.2.3 Collapse rate and threshold in shape space Instantaneous rate and locus.The HCCB collapse rate is γ(t) = η˙qi(t) ˙qj(t)gijq(t)=η κ ZΣ(q(t)) ˙qiui(x;q(t))2w(x;q(t)) dS(x).(161) Hence the locus of collapse (where γ spikes) is the neck Σ( q ), with intensity controlled by the normal component of the instantaneous shape velocity projected onto the moving interface. Geometric threshold law (timing). Collapse of fragment labels occurs when Zt? t0 γ(t)dt =η κ Zt? t0ZΣ(q(t)) ˙qiui2w dS dt ≥Θ(ε),(162) where Θ( ε ) = log (1 /ε )sets the tolerated off–diagonal visibility. This is the neck–localized version of the general threshold law: collapse is triggered by sufficient normal motion of the partition surface, weighted by occupancy in the interfacial region. 5.2.4 Path dependence from noncommuting shape flows Two control directions. Let (Ω , ∆) be two shape coordinates (e.g., elongation Ωand asymmetry ∆). The metric diagonals are gΩΩ =κZΣ u2 Ωw dS, g∆∆ =κZΣ u2 ∆w dS. The order dependence for two protocols that traverse the same small rectangle in opposite orders is governed by the “curl” ∂Ωg∆∆ −∂∆gΩΩ. Shape derivative (Hadamard) calculus. If the surface Σis advected with normal velocity u ( x ), the shape derivative of a surface integral obeys (for vanishing tangential reparametrization) d d=0 ZΣ f dS =ZΣ∂nf+H fu dS, (163) where ∂nf = ∇f·n and H is the mean curvature of Σ. Apply (163) to g∆∆ = κRΣf∆dS with f∆ = u2 ∆w and surface velocity u=uΩ. A direct computation yields ∂Ωg∆∆ =κZΣh2u∆DΩu∆+ (∂nw+Hw)uΩu2 ∆idS, (164) ∂∆gΩΩ =κZΣh2uΩD∆uΩ+ (∂nw+Hw)u∆u2 ΩidS, (165) where DΩu∆ denotes the shape derivative of the normal–velocity field u∆ under the Ω–flow.[ 220 ] Subtracting (165) from (164) gives ∂Ωg∆∆ −∂∆gΩΩ = 2κZΣ [u∆DΩu∆−uΩD∆uΩ]dS +κZΣ (∂nw+Hw)uΩu∆(u∆−uΩ)dS. (166) 39 The leading commutator of shape flows is the antisymmetric part u∆DΩu∆−uΩD∆uΩ=1 2[uΩ, u∆]shape =1 2(uΩ∂nu∆−u∆∂nuΩ) + ··· ,(167) which vanishes only if the two normal–velocity fields commute along n . Equations (166) – (167) exhibit the geometric source of order dependence:noncommuting shape flows (together with curvature/weight gradients on Σ) make ∂Ωg∆∆ −∂∆gΩΩ 6 = 0, producing the area–law difference in collapse exponents for reversed orderings. 5.2.5 Physical consequences and observables Basis. The stabilized basis is the left/right fragment identity defined by the partition Ω L/ Ω R ; it is derived from the projector P(q), not imposed at scission by fiat. Locus. The locus of collapse is the neck Σ( q ): the metric (159) localizes there, and the instantaneous rate (161) is an interfacial quadratic form in the normal shape velocity. Timing. The timing t? satisfies the neck–threshold (162) . Protocols that reach the same endpoint can cross the threshold at different times depending on how they sequence elongation, necking, and asymmetry (via uiand walong the path). Protocol dependence: area law in fission. For two protocols with identical endpoints and duration that traverse a small rectangle in (Ω,∆), the difference in integrated collapse exponents is ΓA−ΓB=η TZZR∂Ωg∆∆ −∂∆gΩΩdΩd∆ + O(k∆k3), with the integrand given by (166) . Hence, order matters: “elongate then bias asymmetry” is not equivalent, for collapse, to “bias asymmetry then elongate,” even if the final scission geometry is the same. Shadow record: scission emissions. With a fast/leaky complement (optical–model coupling to the continuum), the shadow operator E=Q˙ P P induces an output channel with intensity λsc(t) = η κ ZΣ(q(t)) ˙qiui2w dS, (168) so prompt emissions (e.g. scission neutrons, high–energy γ ) provide a direct readout of the interfacial metric speed. Order–reversed protocols with equal endpoints/durations can thus be distinguished by their time–resolved emission records, even when standard flux–based yield predictions would coincide. 5.2.6 Summary HCCB recasts fragment–identity formation as a geometric, neck–localized collapse process. The relevant projector P ( q )is a moving partition of space; its metric (159) is the surface integral of the product of normal shape velocities (weighted by occupancy) over the neck, and the collapse rate is the corresponding quadratic form (161) . Collapse occurs when the integrated interfacial metric speed exceeds a threshold, fixing the timing; the neck Σis the locus; and the left/right partition defines the basis. Crucially, noncommuting shape flows generate a protocol dependence (area law) via (166) , providing a concrete, testable signature beyond the standard description. C. Deterministic vs. probabilistic Idea. In the HCCB account of fission, two distinct layers coexist: • Deterministic (structural). The pointer basis of collapse is fixed by geometry: it is the fragment– number basis determined by the moving partition (neck)—equivalently, by the spectral projectors of a suitable interfacial number operator. The locus (the neck) and the timing (a threshold of integrated metric speed) are likewise fixed by geometry. • Probabilistic (statistical). Given the basis, the specific outcome—e.g. AH =137 vs. 139, AL =95 vs. 97 (post–neutron)—remains stochastic, with weights set by the quantum state supported near scission. These weights are strongly influenced by shell corrections (which shape the collective amplitude and flux) and are modulated by the path–dependent collapse functional Γ. We develop both layers quantitatively. 40 5.3.1 Deterministic layer: why the fragment–number basis is selected Interfacial number operator. Let φ ( x ; q )be a smooth boundary function whose zero–level set Σ( q ) = {φ= 0}is the neck (Sec. V B). Define the smoothed domain projector on the one–body space (P(`) L(q)ψ)(x) = sφ(x;q) `ψ(x), P(`) R=1−P(`) L, with s a smooth step and interface thickness `↓ 0. On Fock space, write the (local) density operator ˆρ(x) = ˆ ψ†(x)ˆ ψ(x)and the left–fragment number operator ˆ A(`) L(q) := ZR3 sφ(x;q) `ˆρ(x)d3x, ˆ A(`) R=Atot −ˆ A(`) L.(169) As `→ 0 + , ˆ A(`) L converges (in the quadratic form sense) to the sharp fragment–number operator that counts nucleons in the region Ω L ( q ) = {φ < 0 } . Let {PA ( q ) } denote the spectral projectors of ˆ AL ( q ) (integer spectrum A= 0,1, . . . , Atot). Jump operator from the moving partition. In the thin–interface limit (Sec. V B), the non–central derivative E=Q˙ P P induces a jump operator that is linear in the boundary–weighted density: L(t)≡√η E(t)≃√η κ ZΣ(q(t)) un(x;q(t)) pw(x;q(t)) ˆρ(x)dS(x).(170) Here un = −˙qi∂iφ/k∇φk is the normal speed of the interface (projected shape velocity) and w is the nonnegative state weight introduced in (159). Define the interfacial number observable ˆ AΣ(t) := ZΣ(q(t)) f(x;t) ˆρ(x)dS(x), f(x;t) := un(x;q(t)) pw(x;q(t)).(171) Then L(t) = √η κ ˆ AΣ(t). Dephasing in the eigenbasis of ˆ AΣ .Consider the (time–local) Lindblad generator Dt [ ρ ] = LρL − 1 2{L2, ρ}with L=√ηκ ˆ AΣ. In the instantaneous eigenbasis {|αi} of ˆ AΣ(t),ˆ AΣ|αi=aα|αi, d dtραβ(t)diss =−ηκ 2aα−aβ2ραβ(t).(172) Thus the dissipator dephases in the eigenbasis of ˆ AΣ ; the off–diagonals decay at an instantaneous rate proportional to the squared eigenvalue difference. From ˆ AΣ to fragment numbers. Because f ( x ; t )in (171) is supported on Σ( q ( t )) and proportional to the normal speed un , ˆ AΣ is an interfacial weighted count: it is (up to a known, positive weight) the generator of variations of ˆ AL under normal interface motion. Indeed, differentiating (169) using s0(φ/`)→` δ(φ)and the coarea formula, d dt ˆ AL(t) = ZΣ(q(t)) ∂tφ k∇φkˆρ dS =−ZΣ(q(t)) unˆρ dS ∝ − ˆ AΣ(t) (up to √w).(173) Therefore, the instantaneous eigenbasis of ˆ AΣ is aligned with the spectral decomposition of ˆ AL ; in the thin–interface and slowly–varying–weight limit, they coincide. Hence (172) is, to excellent approximation, pure dephasing in the fragment–number basis {PA(q(t))}: d dtρAA0(t)diss ≈ − ηκ 2ζ(t)A−A02ρAA0(t), ζ(t) := RΣu2 nw dS RΣw dS .(174) Proposition V.1 (Deterministic pointer basis) . Under the thin–interface and slowly varying weight assumptions, the HCCB dissipator generated by L = √ηκ ˆ AΣ drives any initial state to one that is block diagonal in the fragment–number basis {PA ( q ) } , with the decay of off–diagonals ρAA0 governed by (174) . Consequently, the pointer basis is fixed by geometry (the moving neck): it is the spectral basis of ˆ AL ( q ) (equivalently, ˆ AR). Sketch of proof. Diagonalize ˆ AΣ at each time; use (173) to identify its eigenbasis with that of ˆ AL ; apply (172) and integrate. Gauge changes inside equal–eigenspaces of ˆ AL do not affect the dissipator, establishing pointer–basis invariance.  41 Locus and timing are geometric. The instantaneous rate prefactor in (174) is (up to normalization) exactly the neck–localized metric speed (161) . Therefore: (i) the locus of collapse is the neck Σ( q ), and (ii) the time t? at which all A6 = A0 coherences fall below tolerance ε is set by the geometric threshold (Sec. II D) Zt? 0 η˙qi˙qjgij(q(t)) dt ≥Θ(ε). 5.3.2 Probabilistic layer: why the specific integers remain stochastic Born weights at collapse. Let t? be the (random but sharply concentrated) collapse time determined by the threshold law. The probability to obtain the left–fragment number A at collapse is the Born weight in the pointer basis: ppre(A|q?) = Tr PA(q?)ρ(t− ?), q?:= q(t?).(175) Here ρ ( t− ? )is the pre–collapse state. The post–neutron integer pair ( Apost L, Apost H )is then obtained by folding (175) with prompt–emission distributions conditional on the primary integers and excitation (as in (150)). Shell corrections set a prior. The pre–collapse state ρ ( t− ? )reflects the collective transport on the PES V ( q ): valleys associated with closed shells enhance amplitude where the heavy fragment is near magic Z = 50, N = 82 (snug to 132Sn ), biasing the primary integers toward ( AL, AH ) ≈ (94 , 140). After prompt neutron emission, this maps to post–neutron peaks near (95,137). HCCB as a geometric likelihood. The HCCB mechanism introduces a path–dependent attenuation of coherences and thus a bias in (175) through the timing t? and the locus of collapse. A convenient semiclassical representation writes the weight for a scission channel ξ (e.g. a narrow window of integers around (AL, AH)) as ppre(ξ)∝ZUξW[q(·)] exp−Γ[q(·)]A[q(·)] Dq, (176) where: Uξ collects path segments that reach the scission region associated with ξ ; A [ q ]is the standard (unitary/dissipative) transport weight shaped by V ( q )(hence by shell corrections); and Γ[ q ]is the geometric collapse functional (124) . The factor exp ( − Γ[ q ]) expresses that path histories spending more “metric time” near the neck collapse earlier and inherit the local amplitude distribution at the earlier q? ; paths that delay neck motion collapse later, after further relaxation along the shell valley. Equation (176) is schematic but captures the likelihood role of HCCB. Linear response around a reference protocol. Let A label a narrow mass window and p0 ( A )the probability under a reference protocol q0 ( · ). For a small deformation q = q0 +  δq (fixed endpoints and T), the first variation is d dp(A)=0 =−DδΓ[q0]χAE0+Dδlog A[q0]χAE0,(177) where χA is the indicator functional selecting paths hitting the A –window, and h·i0 denotes averaging with the reference weight. The first term is the geometric contribution from HCCB (via δ Γ, Sec. IV C); the second is the usual dynamical response of the shell–shaped amplitude. Thus even at fixed endpoints and T , reordering segments changes p ( A )by the area–law contribution in δ Γ, a genuinely new, path–dependent effect. 5.3.3 What is determined vs. what remains random Determined by geometry (HCCB). 1. Basis. The collapse basis is the spectral decomposition {PA ( q ) } of the fragment–number operator ˆ AL(q)(Proposition V.1). 2. Locus. Collapse concentrates on the neck Σ(q); the rate density is interfacial:∝u2 nw(Eq. (161)). 3. Timing. Ageometric threshold law Rη˙qi˙qjgij dt ≥Θ(ε)fixes t?. 4. Protocol (order) dependence. For two protocols with identical endpoints/duration, the difference ΓA−ΓBobeys an area law (Sec. IV C), so the record and the induced weights are order sensitive. 48 Geodesic suppression. Among protocols with fixed endpoints/time, those that follow g –geodesics (constant–speed) minimize Γand delay collapse, yielding broader distributions and later scission–emission onsets relative to non–geodesic detours. Reparametrizations at fixed shape track (speed changes) modify Γquadratically; thus, speed shaping is an independent control lever. Curvature tomography. Order–reversed loops on the (∆ , c )plane reconstruct the local curl ∂∆gcc − ∂cg∆∆ from differences in integrated scission counts (Sec. IV A). Mapping this curl over a grid yields a curvature map of the projector bundle, a purely geometric diagnostic inaccessible to standard decoherence models. 5.5.6 Why standard models cannot access these predictions No moving projector, no metric/curvature. Mac–mic transport and TDGCM compute yields from the time–integrated scission flux through a fixed surface (Eqs. (149) , (153) ). They do not endow the fragment–identity assignment with a moving projector, hence they lack: (i) a metric gij that fixes a collapse rate γ = η˙qi˙qjgij , (ii) a geometric threshold for timing, and (iii) a curvature (area law) that imprints order on collapse. Consequently, at fixed endpoints/time and comparable scission flux, these models predict no protocol dependence in widths or time–resolved emissions. HCCB fills this structural gap. 5.5.7 Summary HCCB elevates three aspects of collapse to structural predictions—basis (fragment numbers), locus (neck), and timing (threshold of the integrated metric speed)—and adds a genuinely new ingredient: path dependence. The latter appears as an area law for the collapse exponent and as a hazard–gated sampling of scission flux, producing order–dependent differences in mass–yield widths and in time–resolved scission–emission records, at fixed endpoints and time. These signatures are falsifiable and lie outside the reach of conventional shell +decoherence frameworks precisely because those frameworks do not contain the projector geometry (P, g, F )that HCCB makes explicit. VI. BROADER IMPLICATIONS FOR QUANTUM COMPUTING Thesis. In quantum processors, the relevant subspace is the (possibly dressed) logical or computational manifold, while the complement collects leakage levels, spectator modes, resonators, and the engineered output continuum. As control parameters q ( t )are driven to implement gates and measurements, the corresponding logical projector P ( q )moves on the Grassmannian. The HCCB framework elevates this motion to a first–class design primitive: JHCCB[q(·)] := Γ[q(·)] = ZT 0 η(t) ˙qi(t) ˙qj(t)gijq(t)dt, gij (q) = Tr(∂iP ∂jP),(196) the collapse functional. It quantifies the integrated propensity for non–central leakage (and hence irreversible collapse in the presence of a fast complement). We show how JHCCB can act as a compiler cost, why it must be minimized during coherent gates and maximized during readout, how order–dependent collapse becomes a diagnostic fingerprint of hardware geometry, and how collapse steering can be harvested as a resource. A. HCCB cost as a compiler objective Control manifold, logical projector, and metric. Let U ( q )denote the control–dependent dressing map (isometry) that embeds a fixed logical subspace HL into the physical Hilbert space, and set P(q) = U(q)ΠU(q)†with Πthe fixed projector on HL. Then ∂iP= [ Ai(q), P ] + Ei(q),Ai:= U(∂iU†), Ei:= Q ∂iP P, (197) 49 and the non–central block Eiis the unique seed of HCCB. The metric gij(q) = 2 Re Tr E† iEj=X `∈LX α∈Qh`|∂iP|αihα|∂jP|`i(198) is computable from (dressed) derivative couplings of logical states to leakage channels. Compiler objective (multi–criteria). For a target logical unitary Utar (gate synthesis) and/or a target POVM {Mk}(measurement design), define a composite cost C[q(·)] = λunit Eunit[q(·)] + λmeas Emeas[q(·)] + λHCCB JHCCB[q(·)] + R[q(·)],(199) where Eunit is a terminal–time gate infidelity (e.g. 1 −1 d2 L|Tr ( U† L ( T ) Utar ) |2 with UL the logical evolution), Emeas encodes measurement utility (e.g. readout time to threshold), R enforces bandwidth and amplitude constraints, and λ ’s are design weights. Minimizing (199) drives q ( t )toward geodesic segments of g during unitary gates (reducing JHCCB) and toward high–metric regions during readout (increasing JHCCB). Necessary conditions (Euler–Lagrange / Pontryagin). Ignoring Eunit/meas momentarily, the variation of JHCCB yields the geodesic equations ¨qk+ Γk ij(q) ˙qi˙qj=1 2 ˙η η˙qkwith Γk ij =1 2gk`(∂igj` +∂jgi` −∂`gij),(200) i.e. constant–speed geodesics when η is constant. With the full objective (199) , Pontryagin’s principle supplies a two–point boundary–value problem on the cotangent bundle of M with running Lagrangian L=λHCCBη˙qTg˙q+(penalties); standard shooting or KKT–based interior–point solvers apply. B. Minimize collapse during gates Central vs. non–central control. During coherent gates, we desire Ei≈ 0along the path: the projector should not shear into leakage. The compiler thus prefers control coordinates in which the unitary dressing U(q)varies internally (commutator part [Ai, P]) while Ei≈0. Concretely, Gate design principle: ∂iP≈[Ai, P]⇒gij ≈0on the gate path.(201) Leakage–aware Riemannian geodesics. For a fixed gate time T , the JHCCB –minimizing path between q (0) and q ( T )is a constant–speed geodesic on ( M, g ); for a fixed path, the optimal time allocation is uniform in the metric speed v ( t ) = p˙qTg˙q , since JHCCB = ηRv2dt ≥ηL2 g/T with Lg = Rv dt . Therefore: • Transpilation rule. Map target gates to hardware controls by (i) choosing a coordinate chart with small gin the region traversed, (ii) routing along g–geodesics, (iii) enforcing constant v(t). • Pulse shaping rule. Avoid fast passages through regions where g spikes (“geometric hot spots”): reduce |˙q|locally to keep vbounded. Example: dressed qubit with leakage. Let P ( q )project onto {| 0( q ) i,| 1( q ) i} inside a transmon manifold {|0i,|1i,|2i, . . .}. If a flux control q1= Φ mixes |1i↔|2iwith matrix element h2|∂Φ1i=χ(Φ), then gΦΦ ≈ |h2|∂Φ1i|2+|h2|∂Φ0i|2∼ |χ(Φ)|2, and JHCCB ∝Rη ( t ) |˙ Φ ( t ) |2|χ (Φ( t )) |2dt . Gate compilation thus avoids regions where |χ (Φ) | is large, or slows down there to hold vconstant. C. Maximize collapse during measurement Readout as controlled HCCB. For projective readout in a designated pointer basis {Pµ ( q ) } , we invert the logic: maximize the HCCB rate subject to preserving the pointer labels. This is achieved by steering into regions with large g along directions that preserve the block decomposition of P but produce large Eibetween those blocks and the complement Q. 50 Design rule (QND with maximal HCCB). Let the Hamiltonian commute with the pointer projectors, [ H, Pµ ] = 0 (QND), while the control q modulates the coupling to the fast complement such that Ei = Q∂iPP is maximal in the readout direction. The readout time to reach visibility ε obeys the geometric threshold tread ≈min nt:Zt 0 η(τ) ˙qTg(q(τ)) ˙q dτ ≥Θ(ε)o.(202) Optimizing q ( · )for minimal tread at fixed amplitude/bandwidth is a convex problem in the speed schedule once a path is chosen. Measurement scheduling. Because γ = η v2 is quadratic in speed, increasing the instantaneous speed in high– g regions yields superlinear gains in collapse rate; hence front–loading the trajectory through g–hot–spots (subject to slew limits) shortens readout. D. Order–dependent collapse as a diagnostic fingerprint Protocol pair test. Implement two control loops with identical endpoints and duration that traverse a small rectangle R ⊂ M in opposite orders. Let NT denote the total detected quanta (photons/phonons). HCCB predicts ∆N:= E[N(1→2) T]−E[N(2→1) T]≈¯η TZZR∂q1g22 −∂q2g11dq1dq2,(203) an area law. Measuring ∆ N over a grid reconstructs the curl of the metric diagonals and, covariantly, the projector curvature. This constitutes a geometry tomography of the hardware’s logical projector, sensitive to stray couplings and cross–talk that are invisible to purely Hamiltonian spectroscopy. Noise separation. Because (203) changes sign under order reversal, quasistatic noise contributions (path–independent in leading order) cancel, isolating the geometric signal. Thus order–reversed diagnostics provide a high–contrast calibration tool. E. Collapse steering as a resource Dissipative state preparation. Given a target logical subspace Htar with projector Ptar , design a path q ( t )such that P ( q )flows from the initialization manifold to Ptar while JHCCB is concentrated when P ( q ) overlaps Htar . The environment then purifies the state into Htar by geometric collapse, yielding robust state prep without feedback. Biasing error channels. Because JHCCB penalizes non–central motion, compiler–level shaping can bias the error model (e.g., dephasing vs. amplitude damping) by aligning gate paths with directions of small g for channels one wishes to suppress, while letting g be large in benign directions. This produces error–biased circuits compatible with tailored QEC (e.g., XZZX/twist codes). Measurement–induced entanglement. For multi–qubit projectors P ( q )resolving joint eigen–subspaces (e.g., stabilizers), steering into regions where the stabilizer projector has large g produces fast measurement– induced entanglement (MID) while keeping single–qubit g small—thereby suppressing local collapse. The resource is the geometry of P: one can “aim” collapse at many–body observables. Stabilizer pumping. Alternate central gate segments ( Ei≈ 0) with targeted non–central segments where L∝Q˙ PP implements stabilizer pumping into code space. The pumping rate is set by η˙qTg˙q ; geodesic segments minimize inadvertent leakage between pumps. F. Analytic gradients for geometry–aware compilation First variation. The functional gradient of JHCCB is δJHCCB δqk(t)=−2ηd dtgkj ˙qj+η ∂kgij ˙qi˙qj+ ˙η gkj ˙qj,(204) times a Lagrange multiplier for endpoint constraints. In practice, one computes gij and ∂kgij either (i) from a hardware model (e.g. circuit QED Hamiltonian) by automatic differentiation of P ( q ), or (ii) from data via order–reversal tomography and local finite differences. 51 Second variation and convexity along a path. Along a fixed path direction δq with compact support, the second variation reads δ2JHCCB =ηZ2δ˙qTg δ ˙q+ ˙qT(∂2g[δq, δq]) ˙qdt, which is positive–definite when g varies slowly along the segment, enabling convex line searches in practice. G. Worked micro–examples Single–qubit dispersive readout (cQED). Let q = ( , ∆) be drive amplitude and detuning; P ( q )projects onto the dressed {| 0( q ) i,| 1( q ) i} . Near the cavity pull χ , the metric component g ∝ |∂| 1( q ) i|2 peaks when the drive admixes leakage | 2 i , while g∆∆ peaks near resonance. Readout schedules that sweep detuning first (Path B) then ramp amplitude cross larger g earlier and reach threshold faster than the reverse order (Path A), consistent with the area law. Two–qubit entangling gate with spectator leakage. For a cross–resonance–like gate with control q = (Ω , Φ) (drive and flux), the metric gΦΦ spikes near avoided crossings to spectator states; geodesic compilation routes around those hot spots while keeping unitary phase accumulation intact, decreasing JHCCB without increasing gate time. H. Limits and caveats When geometry is flat. If g is (locally) flat and separable, order–reversal diagnostics vanish to leading order; geometry–aware compilation reduces to constant–speed modulation. Nonzero curvature is the regime where HCCB advantages are largest. Model mismatch. If the P ( q )model is inaccurate, g estimated from data via order–reversal remains reliable. The compiler can close the loop by updating g from fresh diagnostics (geometry–adaptive calibration). Environment scale η .The factor η ( t )sets the irreversibility scale; in near–closed operation η is small and JHCCB is a propensity rather than a realized collapse budget. Nonetheless, minimizing JHCCB still suppresses reversible leakage, improving gate robustness. I. Summary •Compiler cost. The HCCB functional JHCCB = Rη˙qTg˙q dt is an actionable, differentiable objective that quantifies the collapse/ leakage budget of a control schedule. •Gate design. Minimize JHCCB by routing along g –geodesics at constant metric speed and by choosing central control coordinates (Ei≈0) for unitary dressing. •Measurement design. Maximize HCCB subject to preserving the pointer basis to achieve faster, more informative readout; allocate speed where gis large. •Diagnostics. Order–reversal area laws furnish a background–insensitive fingerprint of projector curvature; use them to tomographically calibrate gand detect cross–talk. •Resource. Collapse steering—via the geometry of P —enables dissipative state prep, measurement– induced entanglement, and error–bias shaping; it is a resource, not merely a nuisance. The upshot is a geometry–aware stack for quantum control: design gates to avoid non–central projector motion and design measurements to exploit it. The same mathematical object—the HCCB cost— organizes both tasks and yields new diagnostics and performance guarantees absent from standard, purely Hamiltonian compilations. 52 VII. DISCUSSION Synopsis. We have formulated collapse as a geometric threshold phenomenon for a moving projector P ( q ), with instantaneous rate γ ( t ) = η˙qi˙qjgij ( q ( t )) and integrated collapse functional JHCCB = Γ[ q ( · )] = RT 0γ ( t ) dt . The core predictions—basis (fragment identity, stabilizers), locus (neck surfaces, readout hot–spots), timing (threshold law), and order dependence (area law)—emerge from the non–central derivatives Ei := Q ∂iP P and the projector geometry ( g, F ). We now place these results in a broader conceptual and practical context. A. Geometric obstructions and categorical non-commutativity From paths to 2–cells. Let Path ( M )be the 2–category whose objects are points q∈M , whose 1–morphisms are piecewise–smooth paths γ : [0 , 1] →M with fixed endpoints, and whose 2–morphisms are endpoint–fixed homotopies H : [0 , 1] 2→M (“bigons”). Let Proj ( H )be the 2–category whose objects are orthogonal projectors on H , whose 1–morphisms are partial isometries intertwining ranges, and whose 2–morphisms are intertwiners modulo phases. Projector functor and connection. The assignment Φ : Path(M)−→ Proj(H), q 7→ P(q), γ 7→ PexpRγ∇, with ∇ the projector connection one–form (see below), is a functor up to 2–isomorphism. Its failure to be strict is measured by a curvature 2–cell determined by Ai:= P ∂iP Q −Q ∂iP P and Fij := P[∂iP, ∂jP]P=E† iEj−E† jEi. The “triangle–level” (central) pieces [ Ai, P ]provide parallel transport internal to Ran P (Berry/Wilczek– Zee), while the non–central blocks Ei create 2–categorical obstructions: the square formed by two infinitesimal displacements fails to commute in Proj(H)by an amount controlled by Fij. Proposition VII.1 (Square non-commutativity ⇒ area law) . Let γ1→2 and γ2→1 be two orderings of small displacements δq1, δq2 forming a rectangle R . Then, for any time budget T shared by both orderings, Γ[γ1→2]−Γ[γ2→1] = η TZZR∂q1g22 −∂q2g11dq1dq2=η TZZR Ξ(F;E)dS +O(kRk3/2), where Ξ( F ; E )is a gauge–invariant scalar contraction built from F12 and the E –blocks. In particular, if F12 ≡0in the region, then the rightmost integral vanishes to leading order. Sketch. The first equality is the metric curl result proved in Sec. IV C. The second follows by the projector calculus identity ∂igjj −∂jgii = 4 Re Tr E† j ( ∂iEj ) −E† i ( ∂jEi )  combined with ∂iEj−∂jEi = EjE† i−EiE† j (a consequence of P2=P), which exhibits the F–dependence through E† iEj−E† jEi=Fij. Central vs. non-central coherence. Central transport ( Ei =0) furnishes a flat 2–functor at the square level: order only accrues phases; no collapse. Non–central transport ( Ei6 =0) introduces 2–cells that are not central: the square witnesses a structural obstruction. In open dynamics with a fast complement, these obstructions become Lindblad jump channels Li∝Ei ; their quadratic form is precisely the HCCB rate γ(t) = η˙qi˙qjgij. B. Shadow as structural record vs. standard environment entanglement Two notions of “shadow.” In HCCB, the shadow vector is E ( t ) |ψi = Q˙ P P|ψi , an explicitly computable kinematic object whose squared norm determines the jump intensity and whose phase (in diffusive unravelings) encodes the direction of motion. In standard decoherence, the “shadow” is the (typically inaccessible) environmental conditional |a ( t ) i ; decoherence is governed by their overlaps (Loschmidt echo) or by bath spectra via filter functions. Theorem VII.2 (Order–reversal null for fixed–channel decoherence) . Consider a Markovian master equation with fixed Lindblad operators {Lα} and constant rates {γα} : ˙ρ = −i ~ [ H ( t ) , ρ ] + Pαγα ( LαρL† α− 1 2{L† αLα, ρ} ). Let HA ( t ) , HB ( t )be two Hamiltonian schedules with the same endpoints and duration T . 53 Then the expected total jump counts for channel α , E [ Nα,T ] = γαRT 0Tr ( L† αLαρ ( t )) dt , are equal for A and B to leading order in the interaction picture, provided the prepared state and Lα are the same and H acts only within the relevant subspace. Sketch. Move to the interaction picture with respect to H ( t ); the dissipator is unchanged in form but with Lα ( t ) = U† ( t ) LαU ( t ). If Lα acts trivially on the Hamiltonian control sector (or commutes with it), then Tr ( L† αLαρ )is invariant to path order; the integral depends on duration and preparation, not on order. In HCCB, by contrast, L∝E=Q˙ P P and does depend on the path. Operational difference. Order–reversal protocols provide a signed diagnostic in HCCB (area law, Sec. IV C); the sign flips when the order flips. In standard fixed–channel decoherence, the leading–order difference vanishes; any residual effect is second–order in path–dependent dressing of Lα and does not display a robust signed area law. C. Verification strategies 7.3.1 Path–order tests (A vs. B) Two–segment test. Choose two control coordinates ( q1, q2 )with a small rectangle R of sides (∆ q1, ∆ q2 ) and duration T. Execute: A: (q1→q1+ ∆q1)then (q2→q2+ ∆q2), B : (q2→q2+ ∆q2)then (q1→q1+ ∆q1). Measure total counts NTin the HCCB channel. Prediction (Sec. VI D): ∆N=E[N(A) T]−E[N(B) T]≈¯η T(∆q1∆q2)∂q1g22 −∂q2g11q0 . Null: swapping order reverses the sign; swapping the rectangle orientation reverses the sign again; cancelling both yields zero. Power estimate. For Poissonian counts with means µA,B , a z –test detects ∆ µ at significance α and power 1−βin nrepeats if n&(z1−α/2+z1−β)2(µA+µB) (µA−µB)2. Since µA,B ∼ Γ A,B , the required n scales as 1 / ∆Γ 2 ; hence larger rectangles (within linearity) and higher ¯η/speed reduce sample complexity. 7.3.2 Scaling with speed Quadratic–in–speed law. For a fixed path shape, rescale time t7→ t/s with s > 0(speedup by factor 1/s). Then Γ7→ Γ0=1 sΓ: Γ0[q(·/s)] = ZT/s 0 η(dq dt0)Tg(dq dt0)dt0=1 sZT 0 η˙qTg˙q dt =Γ s. Thus counts scale linearly with speedups for fixed path. Within bounds ( η constant), this provides a clean check that the measured intensity is geometric (∝v2), not a fixed–rate bath artifact. 7.3.3 Scaling with area Area law. For small rectangles, ∆Γ ∝Area ( R ) = ∆ q1 ∆ q2 . Doubling either side doubles ∆Γ; reversing orientation flips the sign. Higher–order corrections scale as O ( kRk3/2 )and can be bounded by curvature derivatives. 7.3.4 Geometry tomography Curl map. Tile a region by small rectangles; measure ∆ N for order–reversed traversals; infer local values of ∂q1g22 −∂q2g11 ; reconstruct a curl map. Combine with single–axis ramps ( λ ( t ) = η˙q2 igii ) to reconstruct gii; together, these determine gup to a gauge inside Ran P. 54 D. Potential impact across domains 7.4.1 Fission physics Basis/locus/timing & path dependence. HCCB formalizes fragment–identity collapse as neck–localized, with a geometric threshold law for timing and an intrinsic pointer basis (fragment numbers). Crucially, it predicts order–dependent yield widths and time–resolved scission emission records for protocols with identical endpoints/time. Experimental levers include pre–scission excitation sequences (e.g., deformation before asymmetry vs. after), controlled neutron–induced fission with shaped energy deposition, and electromagnetic biasing of asymmetry prior to necking. Diagnostic observables. (i) Early–time scission neutrons/ γ : Path B (asymmetry → neck) front–loads intensity (Sec. V D). (ii) Mass–yield widths: narrower heavy peak for Path B at fixed endpoints. (iii) TKE–mass correlations: slightly reduced TKE at given split for earlier collapse (thicker neck at threshold). These are not implied by standard shell +flux models and thus serve as falsifiable HCCB signatures. 7.4.2 Quantum processors Compiler metric and diagnostics. Treat JHCCB as a cost for gates and a budget for readout (Sec. VI). Geometry–aware compilation follows g –geodesics during gates (minimizing JHCCB ) and traverses high– g regions during measurement (maximizing it). Order–reversal counts provide a robust fingerprint of projector curvature, complementing Hamiltonian spectroscopy and revealing cross–talk to leakage channels. Resource view. Collapse steering can be harvested for measurement–induced entanglement, stabilizer pumping, and error–bias tailoring; it turns a nuisance (leakage) into an engineered resource. 7.4.3 Biological decision processes Mapping. Let q be external/internal cues (ligands, stresses, metabolic signals), and let P ( q )project onto a phenotypic or regulatory subspace (e.g., transcriptional network modes associated with a fate). The complement contains alternative fates or transient programs. As cues vary, the relevant subspace moves; non–central shear E=Q˙ P P quantifies leakage into alternative programs. Predictions. (i) Order dependence: exposure A → B vs. B → A yields different fate distributions at fixed dose/time (area law on cue space). (ii) Speed scaling: faster cue changes increase commitment rate quadratically in speed along high–susceptibility directions ( g hot–spots). (iii) Locus: decision points (commitment) concentrate near interfaces in regulatory landscapes (analogues of necks), testable by time–resolved single–cell trajectories with controlled stimulus ordering. E. Limitations, assumptions, and open questions Separation of timescales. Our reduction relies on a fast, relaxing complement (Liouvillian gap κ > 0) and slow controls (TCL/adiabatic assumptions). In regimes where Q retains coherence on gate timescales, E6 = 0 still produces reversible leakage, but the irreversible interpretation of Γas a collapse budget must be qualified. State dependence. While γ ( t )is state–independent, observed counts involve Tr ( E†E ρP ); protocol comparisons should use identical preparations or average over them. Our diagnostic formulas either assume this or explicitly model preparation variability. Higher–order geometry. We emphasized metric g and curvature F ; more refined signatures may involve higher covariant derivatives (“geometric jerk”) and multi–parameter holonomies in nonabelian settings (rank r > 1). Extending area laws beyond leading order is an open technical problem with potentially richer experimental structure. Categorical generalization. We provided a 2–categorical reading (squares). The “higher categorical coherence breakdown” (HCCB) suggests that obstructions at tetrahedra and higher cells (noncentral 3–cocycles etc.) may classify families of collapse channels beyond Ei—a subject for future work. 55 F. Concluding perspective HCCB reframes collapse as a geometric and structural effect of moving projectors. The shadow is not merely a statistical afterthought of tracing out an environment; it is a computable kinematic record set by ( P, ˙ P ), with a rate fixed by the projector metric and with protocol memory quantified by curvature. This viewpoint unifies diverse arenas—nuclear fission, quantum information processing, and biological decision making—under a single mathematical umbrella and yields concrete, falsifiable predictions: 1. Basis/locus/timing become derivable from geometry. 2. Order dependence appears as an area law for collapse. 3. Collapse steering becomes an optimizable resource. The chief experimental task is clear: perform path–order tests with speed/area scaling and read out the shadow by counting emissions or outcomes. A positive result would elevate projector geometry to a practical design principle across physics and biology; a null (modulo control confounds) would impose sharp constraints on where non–central obstructions can (and cannot) arise. Either way, the approach provides a new lens on how dynamics, structure, and measurement entwine. VIII. CONCLUSION Statement of the thesis. We have argued that collapse is a structural, geometric inevitability whenever the relevant subspace of a quantum system is implemented by a moving projector P ( q )whose non–central derivatives Ei:= Q ∂iP P do not vanish. The unique quadratic invariant built from these derivatives, gij(q) = Tr ∂iP ∂jP= 2 Re Tr E† iEj,(205) defines a Riemannian metric on the control manifold, and the collapse rate is the metric speed γ(t) = η˙qi(t) ˙qj(t)gijq(t),(206) so that the integrated collapse functional Γ[q(·)] = ZT 0 γ(t)dt (207) plays the role of a geometric action. Collapse to a pointer basis occurs once Γexceeds a threshold set by the desired visibility Θ( ε ) = log (1 /ε ). The order in which controls are applied matters: for two protocols with the same endpoints and duration, ΓA−ΓB≈η TZZR∂q1g22 −∂q2g11dq1dq2, an area law whose geometric source is the projector curvature Fij = P [ ∂iP, ∂jP ] P = E† iEj−E† jEi . In short, the non–central derivatives Ei are the seed of collapse; the metric g sets its rate; and the curvature Fimprints history on the record. A. What HCCB predicts beyond standard quantum mechanics Basis (what becomes definite). Standard decoherence models typically assume a measurement channel and thus a basis. Here, the basis is derived: the jump operator L = √η E dephases in the eigenbasis of the observable generated by interfacial motion of P , which—in the cases studied—coincides with the spectral projectors of the structurally relevant quantity (e.g., fragment numbers in fission; stabilizers or logical labels in quantum processors). Locus (where collapse concentrates). The rate density is localized where the projector changes non– centrally. In fission this is the neck surface; in readout hardware it is the region of strong hybridization with the output continuum; in biological decision processes it is the interface between regulatory modules. The coarea reduction shows that g collapses to a surface integral over the interface, making these locations hot–spots of collapse. 56 Timing (when collapse happens). The threshold law Zt? 0 η˙qi˙qjgij(q(t)) dt ≥Θ(ε) turns timing into a geometric problem. Among paths with fixed endpoints and duration, constant–speed geodesics of gminimize Γand thus delay collapse; detours that increase the g–length advance it. Path dependence (how history matters). Standard fixed–channel GKSL models are, to leading order, path independent at fixed endpoints and time. In HCCB, Γis a functional of the path; order–reversed traversals of the same rectangle differ by a signed area determined by ∂g (equivalently, by F ). This yields diagnostic and control consequences that are inaccessible to conventional decoherence models. B. Demonstrated case: nuclear fission Geometric neck metric and threshold. Modeling the left/right projector by a level–set partition across the neck, shape calculus gives a neck–localized metric gij(q) = κZΣ(q) ui(x;q)uj(x;q)w(x;q)dS, ui=−∂iφ k∇φk, so that the instantaneous rate is γ(t) = η κ ZΣ(q(t))˙qiui2w dS. For neck radius c , gcc ∝ 1 / ( c + c0 )diverges as the waist thins; the collapse threshold is therefore inevitably reached near scission. Predictive leverage beyond shell corrections. Shell corrections shape the prior amplitude along the path (hence the heavy–fragment bias near magic numbers), while HCCB fixes when that amplitude is sampled and where collapse concentrates. Two protocols with identical endpoints/time—(A) elongate then neck and (B) bias asymmetry then neck—have different integrated exponents Γand thus different collapse times: Path B collapses earlier at a thicker neck, predicting front–loaded scission emissions and narrower heavy–fragment widths than Path A. These order–dependent signatures do not follow from standard “flux through scission surface” rules and constitute falsifiable HCCB predictions. C. New control paradigms for quantum computing Collapse functional as compiler cost. In quantum processors the same geometry appears with P ( q ) projecting onto a dressed logical manifold. The functional JHCCB[q(·)] = ZT 0 η(t) ˙qTg(q(t)) ˙q dt is a differentiable cost for gates (to be minimized) and a budget for readout (to be maximized subject to QND constraints). Geodesic routing and constant metric speed suppress collapse during coherent operations; targeted excursions into high– g regions accelerate measurement. Order–reversal area laws furnish geometry tomography of P , exposing cross–talk invisible to Hamiltonian spectroscopy. Collapse steering becomes a resource: for dissipative state preparation, stabilizer pumping, measurement–induced entanglement, and error–bias tailoring. D. Beyond physics: biological decision processes Mapping cues to a control vector q and phenotypic/regulatory modes to a projector P ( q ), HCCB predicts order and speed effects in fate commitment: A → B stimuli vs. B → A produce different commitment times and distribution widths at fixed total dose/time (an area law on cue space), with commitment localized at interfaces (network “necks”). These are structurally the same invariants (P, g, F ). 57 E. Verification playbook and nulls Order tests and scalings. (i) Order reversal: traverse a small rectangle in opposite orders; the difference in total counts (photons/phonons, scission quanta) scales with the signed area and flips sign under order swap. (ii) Speed scaling: at fixed path, counts scale linearly with the inverse duration (Γ 7→ Γ /s under a time rescaling by s ), reflecting the v2 rate law. (iii) Geodesic vs. detour: at equal duration, detours with larger g –length yield earlier thresholds. Nulls occur in locally flat/separable–metric regions ( ∂g =0) or when shape flows commute (vanishing curvature). F. Limitations and open directions Assumptions. The reduction to a rate γ = η˙qTg˙q presumes a fast, relaxing complement (Markov/TCL regime) and slow control; outside this regime, memory kernels dress η and E , but the geometric dependence persists. State dependence enters only through Tr ( E†E ρP ); protocol comparisons should match preparations or average over them. Higher coherence breakdown. We identified the 2–cell obstruction (curvature F) responsible for area laws and indicated 3–cell (Bianchi–type) terms that can appear beyond the adiabatic Markovian limit. Characterizing and measuring such higher obstructions—both mathematically and operationally—remains a fertile direction. Bridges to other theories. The same categorical viewpoint suggests formal functors to gravitational settings (where higher coherence breakdown parallels geometric curvature) and to mesoscale systems (neuronal, biochemical). Establishing these bridges rigorously is an open program. G. Key equations (cheat sheet) Shadow operator: E(t) = Q˙ P P, kE|ψik2=hψ|E†E|ψi. Metric: gij = Tr(∂iP ∂jP) = 2 Re Tr(E† iEj). Rate and action: γ(t) = η˙qi˙qjgij(q(t)),Γ[q] = ZT 0 γ(t)dt. Threshold: Γ[0→t?]≥Θ(ε) = log(1/ε). Area law (small loops): ΓA−ΓB≃η TZZR (∂q1g22 −∂q2g11)dq1dq2. Curvature: Fij =P[∂iP, ∂jP]P=E† iEj−E† jEi. H. Closing perspective The higher–categorical coherence breakdown (HCCB) program replaces ad hoc collapse stories with a single geometric mechanism grounded in the motion of projectors. It delivers concrete predictions that standard quantum mechanics, in its fixed–channel open–system form, does not: which observables become definite (basis), where collapse concentrates (locus), when it occurs (timing), and how it records history (path dependence). The nuclear fission case shows added predictive power beyond shell corrections, and the quantum–computing analysis turns collapse from a nuisance into a design resource. The same invariants (P, g, F)give a common language across domains, from nuclei to processors to cells. The roadmap is empirical and sharp: perform path–order tests, verify speed/area scalings, and map curvature via order–reversed loops. A positive outcome would establish projector geometry as a new organizing principle for dynamics and measurement; a negative outcome would set stringent limits on non–central obstructions. Either way, HCCB reframes long–standing questions about why and how collapse happens into a tractable, quantitative geometry—opening avenues for control, computation, and understanding across the sciences. 64 Notation index (Appendices) P(q)Relevant projector (moving subspace) Q=1−PComplement projector Ei=Q ∂iP P Non-central derivative block (leakage seed) gij = Tr(∂iP ∂jP)Projector (Grassmannian) metric Fij =P[∂iP, ∂jP]PProjector curvature (2-cell obstruction) γ(t) = η˙qi˙qjgij HCCB instantaneous collapse rate Γ = Rγ dt Integrated collapse functional (geometric action) Θ(ε) = log(1/ε)Visibility threshold parameter ˆ AΣInterfacial (pointer) observable ∆aαβ Eigenvalue difference of ˆ AΣ [1] M. Born, “Zur Quantenmechanik der Stoßvorgänge,” Zeitschrift für Physik 37, 863–867 (1926). [2] J. von Neumann, Mathematische Grundlagen der Quantenmechanik (Springer, Berlin, 1932). [3] V. Gorini, A. Kossakowski, and E. C. G. Sudarshan, “Completely positive dynamical semigroups of N -level systems,” J. Math. Phys. 17, 821–825 (1976). [4] G. Lindblad, “On the generators of quantum dynamical semigroups,” Commun. Math. Phys. 48 , 119–130 (1976). [5] H.-P. Breuer and F. Petruccione, The Theory of Open Quantum Systems (Oxford University Press, 2002). [6] A. G. Redfield, “On the theory of relaxation processes,” IBM J. Res. Dev. 1, 19–31 (1957). [7] W. H. Zurek, “Decoherence, einselection, and the quantum origins of the classical,” Rev. Mod. Phys. 75 , 715–775 (2003). [8] E. Joos, H. D. Zeh, C. Kiefer, D. Giulini, J. Kupsch, and I.-O. Stamatescu, Decoherence and the Appearance of a Classical World in Quantum Theory, 2nd ed. (Springer, 2003). [9] M. Schlosshauer, Decoherence and the Quantum-to-Classical Transition (Springer, 2007). [10] H. J. Carmichael, An Open Systems Approach to Quantum Optics (Springer, 1993). [11] H. M. Wiseman and G. J. Milburn, Quantum Measurement and Control (Cambridge University Press, 2010). [12] T. Kato, “On the adiabatic theorem of quantum mechanics,” J. Phys. Soc. Jpn. 5, 435–439 (1950). [13] J. P. Provost and G. Vallée, “Riemannian structure on manifolds of quantum states,” Commun. Math. Phys. 76, 289–301 (1980). [14] M. V. Berry, “Quantal phase factors accompanying adiabatic changes,” Proc. R. Soc. A 392 , 45–57 (1984). [15] F. Wilczek and A. Zee, “Appearance of gauge structure in simple dynamical systems,” Phys. Rev. Lett. 52 , 2111–2114 (1984). [16] J. Anandan and Y. Aharonov, “Geometry of quantum evolution,” Phys. Rev. Lett. 65, 1697–1700 (1990). [17] R. Resta, “The insulating state of matter: A geometrical theory,” Eur. Phys. J. B 79, 121–137 (2011). [18] M. Kolodrubetz, D. Sels, P. Mehta, and A. Polkovnikov, “Geometry and non-adiabatic response in quantum and classical systems,” Phys. Rep. 697, 1–87 (2017). [19] S. Nakajima, “On quantum theory of transport phenomena,” Prog. Theor. Phys. 20, 948–959 (1958). [20] R. Zwanzig, “Ensemble method in the theory of irreversibility,” J. Chem. Phys. 33, 1338–1341 (1960). [21] E. B. Davies, Quantum Theory of Open Systems (Academic Press, 1976). [22] N. Bohr and J. A. Wheeler, “The mechanism of nuclear fission,” Phys. Rev. 56, 426–450 (1939). [23] U. Brosa, S. Grossmann, and A. Müller, “Nuclear fission process,” Phys. Rep. 197, 167–262 (1990). [24] J. Randrup and P. Möller, “Brownian shape motion on five-dimensional potential-energy surfaces: Nuclear fission-fragment mass distributions,” Phys. Rev. Lett. 106, 132503 (2011). [25] V. Gorini, A. Kossakowski, and E. C. G. Sudarshan, “Completely positive dynamical semigroups of N -level systems,” J. Math. Phys. 17, 821–825 (1976). [26] G. Lindblad, “On the generators of quantum dynamical semigroups,” Commun. Math. Phys. 48 , 119–130 (1976). [27] H.-P. Breuer and F. Petruccione, The Theory of Open Quantum Systems (Oxford University Press, 2002). [28] W. H. Zurek, “Decoherence, einselection, and the quantum origins of the classical,” Rev. Mod. Phys. 75 , 715–775 (2003). [29] E. Joos, H. D. Zeh, C. Kiefer, D. Giulini, J. Kupsch, and I.-O. Stamatescu, Decoherence and the Appearance of a Classical World in Quantum Theory, 2nd ed. (Springer, 2003). [30] M. Schlosshauer, Decoherence and the Quantum-to-Classical Transition (Springer, 2007). [31] H. J. Carmichael, An Open Systems Approach to Quantum Optics (Springer, 1993). [32] H. M. Wiseman and G. J. Milburn, Quantum Measurement and Control (Cambridge University Press, 2010). [33] K. Jacobs and D. A. Steck, “A straightforward introduction to continuous quantum measurement,” Contemp. Phys. 47, 279–303 (2006). [34] W. H. Zurek, “Preferred states, predictability, classicality and the environment-induced decoherence,” Prog. Theor. Phys. 89, 281–312 (1993). [35] R. B. Griffiths, “Consistent histories and the interpretation of quantum mechanics,” J. Stat. Phys. 36 , 219–272 (1984). 65 [36] M. Gell-Mann and J. B. Hartle, “Quantum mechanics in the light of quantum cosmology,” in Complexity, Entropy and the Physics of Information, ed. W. H. Zurek (Addison-Wesley, 1990), pp. 425–458. [37] M. Gell-Mann and J. B. Hartle, “Classical equations for quantum systems,” Phys. Rev. D 47 , 3345–3382 (1993). [38] R. Omnès, The Interpretation of Quantum Mechanics (Princeton University Press, 1994). [39] G. C. Ghirardi, A. Rimini, and T. Weber, “Unified dynamics for microscopic and macroscopic systems,” Phys. Rev. D 34, 470–491 (1986). [40] P. Pearle, “Combining stochastic dynamical state-vector reduction with spontaneous localization,” Phys. Rev. A 39, 2277–2289 (1989). [41] A. Bassi, K. Lochan, S. Satin, T. P. Singh, and H. Ulbricht, “Models of wave-function collapse, underlying theories, and experimental tests,” Rev. Mod. Phys. 85, 471–527 (2013). [42] A. O. Caldeira and A. J. Leggett, “Quantum tunnelling in a dissipative system,” Ann. Phys. 149 , 374–456 (1983). [43] C. Zener, “Non-adiabatic crossing of energy levels,” Proc. R. Soc. A 137, 696–702 (1932). [44] T. Kato, “On the adiabatic theorem of quantum mechanics,” J. Phys. Soc. Jpn. 5, 435–439 (1950). [45] M. S. Sarandy and D. A. Lidar, “Adiabatic approximation in open quantum systems,” Phys. Rev. A 73 , 062101 (2006). [46] J. P. Provost and G. Vallée, “Riemannian structure on manifolds of quantum states,” Commun. Math. Phys. 76, 289–301 (1980). [47] M. V. Berry, “Quantal phase factors accompanying adiabatic changes,” Proc. R. Soc. A 392 , 45–57 (1984). [48] F. Wilczek and A. Zee, “Appearance of gauge structure in simple dynamical systems,” Phys. Rev. Lett. 52 , 2111–2114 (1984). [49] J. Anandan and Y. Aharonov, “Geometry of quantum evolution,” Phys. Rev. Lett. 65, 1697–1700 (1990). [50] M. Kolodrubetz, D. Sels, P. Mehta, and A. Polkovnikov, “Geometry and non-adiabatic response in quantum and classical systems,” Phys. Rep. 697, 1–87 (2017). [51] D. Xiao, M.-C. Chang, and Q. Niu, “Berry phase effects on electronic properties,” Rev. Mod. Phys. 82 , 1959–2007 (2010). [52] I. Bengtsson and K. Życzkowski, Geometry of Quantum States (Cambridge University Press, 2006). [53] T. Kato, “On the adiabatic theorem of quantum mechanics,” J. Phys. Soc. Jpn. 5, 435–439 (1950). [54] J. E. Avron and A. Elgart, “Adiabatic theorem without a gap condition,” Commun. Math. Phys. 203 , 445–463 (1999). [55] E. Brion, L. H. Pedersen, and K. Mølmer, “Adiabatic elimination in a lambda system,” J. Phys. A 40 , 1033–1043 (2007). [56] F. Reiter and A. S. Sørensen, “Effective operator formalism for open quantum systems,” Phys. Rev. A 85 , 032111 (2012). [57] D. Sels and A. Polkovnikov, “Minimizing irreversible losses in quantum systems by local counterdiabatic driving,” Proc. Natl. Acad. Sci. USA 114, E3909–E3916 (2017). [58] S. Nakajima, “On quantum theory of transport phenomena,” Prog. Theor. Phys. 20, 948–959 (1958). [59] R. Zwanzig, “Ensemble method in the theory of irreversibility,” J. Chem. Phys. 33, 1338–1341 (1960). [60] E. B. Davies, Quantum Theory of Open Systems (Academic Press, 1976). [61] H.-P. Breuer and F. Petruccione, The Theory of Open Quantum Systems (Oxford University Press, 2002). [62] Á. Rivas, S. F. Huelga, and M. B. Plenio, “Quantum non-Markovianity: Characterization, quantification and detection,” Rep. Prog. Phys. 77, 094001 (2014). [63] P. Zanardi and M. Rasetti, “Holonomic quantum computation,” Phys. Lett. A 264, 94–99 (1999). [64] F. Ticozzi and L. Viola, “Quantum Markovian subsystems: Invariance, attractivity, and control,” IEEE Trans. Autom. Control 53, 2048–2063 (2008). [65] H. J. Carmichael, An Open Systems Approach to Quantum Optics (Springer, 1993). [66] H. M. Wiseman and G. J. Milburn, Quantum Measurement and Control (Cambridge University Press, 2010). [67] K. Jacobs and D. A. Steck, “A straightforward introduction to continuous quantum measurement,” Contemp. Phys. 47, 279–303 (2006). [68] J. Dalibard, Y. Castin, and K. Mølmer, “Wave-function approach to dissipative processes in quantum optics,” Phys. Rev. Lett. 68, 580–583 (1992). [69] M. V. Berry, “Quantal phase factors accompanying adiabatic changes,” Proc. R. Soc. A 392 , 45–57 (1984). [70] F. Wilczek and A. Zee, “Appearance of gauge structure in simple dynamical systems,” Phys. Rev. Lett. 52 , 2111–2114 (1984). [71] J. P. Provost and G. Vallée, “Riemannian structure on manifolds of quantum states,” Commun. Math. Phys. 76, 289–301 (1980). [72] M. Kolodrubetz, D. Sels, P. Mehta, and A. Polkovnikov, “Geometry and non-adiabatic response in quantum and classical systems,” Phys. Rep. 697, 1–87 (2017). [73] H.-P. Breuer and F. Petruccione, The Theory of Open Quantum Systems (Oxford University Press, 2002). [74] F. Reiter and A. S. Sørensen, “Effective operator formalism for open quantum systems,” Phys. Rev. A 85 , 032111 (2012). [75] F. Ticozzi and L. Viola, “Quantum Markovian subsystems: Invariance, attractivity, and control,” IEEE Trans. Autom. Control 53, 2048–2063 (2008). [76] I. Bengtsson and K. Życzkowski, Geometry of Quantum States (Cambridge University Press, 2006). [77] W. H. Zurek, “Decoherence, einselection, and the quantum origins of the classical,” Rev. Mod. Phys. 75 , 66 715–775 (2003). [78] M. Schlosshauer, Decoherence and the Quantum-to-Classical Transition (Springer, 2007). [79] D. Sels and A. Polkovnikov, “Minimizing irreversible losses in quantum systems by local counterdiabatic driving,” Proc. Natl. Acad. Sci. USA 114, E3909–E3916 (2017). [80] M. Demirplak and S. A. Rice, “Adiabatic population transfer with control fields,” J. Phys. Chem. A 107 , 9937–9945 (2003). [81] M. V. Berry, “Transitionless quantum driving,” J. Phys. A 42, 365303 (2009). [82] J. Anandan and Y. Aharonov, “Geometry of quantum evolution,” Phys. Rev. Lett. 65, 1697–1700 (1990). [83] S. Deffner and S. Campbell, “Quantum speed limits: from Heisenberg’s uncertainty principle to optimal quantum control,” J. Phys. A 50, 453001 (2017). [84] M. S. Sarandy and D. A. Lidar, “Adiabatic approximation in open quantum systems,” Phys. Rev. A 73 , 062101 (2006). [85] T. Kato, “On the adiabatic theorem of quantum mechanics,” J. Phys. Soc. Jpn. 5, 435–439 (1950). [86] J. E. Avron and A. Elgart, “Adiabatic theorem without a gap condition,” Commun. Math. Phys. 203 , 445–463 (1999). [87] H. J. Carmichael, An Open Systems Approach to Quantum Optics (Springer, 1993). [88] H. M. Wiseman and G. J. Milburn, Quantum Measurement and Control (Cambridge University Press, 2010). [89] K. Jacobs and D. A. Steck, “A straightforward introduction to continuous quantum measurement,” Contemp. Phys. 47, 279–303 (2006). [90] M. A. Nielsen and I. L. Chuang, Quantum Computation and Quantum Information, 10th Anniversary Ed. (Cambridge University Press, 2010). [91] S. Nakajima, “On quantum theory of transport phenomena,” Prog. Theor. Phys. 20, 948–959 (1958). [92] R. Zwanzig, “Ensemble method in the theory of irreversibility,” J. Chem. Phys. 33, 1338–1341 (1960). [93] E. B. Davies, Quantum Theory of Open Systems (Academic Press, 1976). [94] H.-P. Breuer and F. Petruccione, The Theory of Open Quantum Systems (Oxford University Press, 2002). [95] M. S. Sarandy and D. A. Lidar, “Adiabatic approximation in open quantum systems,” Phys. Rev. A 73 , 062101 (2006). [96] J. P. Provost and G. Vallée, “Riemannian structure on manifolds of quantum states,” Commun. Math. Phys. 76, 289–301 (1980). [97] M. V. Berry, “Quantal phase factors accompanying adiabatic changes,” Proc. R. Soc. A 392 , 45–57 (1984). [98] F. Wilczek and A. Zee, “Appearance of gauge structure in simple dynamical systems,” Phys. Rev. Lett. 52 , 2111–2114 (1984). [99] M. Kolodrubetz, D. Sels, P. Mehta, and A. Polkovnikov, “Geometry and non-adiabatic response in quantum and classical systems,” Phys. Rep. 697, 1–87 (2017). [100] J. Anandan and Y. Aharonov, “Geometry of quantum evolution,” Phys. Rev. Lett. 65, 1697–1700 (1990). [101] I. Bengtsson and K. Życzkowski, Geometry of Quantum States (Cambridge University Press, 2006). [102] H. J. Carmichael, An Open Systems Approach to Quantum Optics (Springer, 1993). [103] H. M. Wiseman and G. J. Milburn, Quantum Measurement and Control (Cambridge University Press, 2010). [104] K. Jacobs and D. A. Steck, “A straightforward introduction to continuous quantum measurement,” Contemp. Phys. 47, 279–303 (2006). [105] J. Gambetta, A. Blais, M. Boissonneault, A. A. Houck, D. I. Schuster, and S. M. Girvin, “Quantum trajectory approach to circuit QED: Quantum jumps and the Zeno effect,” Phys. Rev. A 77, 012112 (2008). [106] A. Blais, A. L. Grimsmo, S. M. Girvin, and A. Wallraff, “Circuit quantum electrodynamics,” Rev. Mod. Phys. 93, 025005 (2021). [107] N. Bohr and J. A. Wheeler, “The mechanism of nuclear fission,” Phys. Rev. 56, 426–450 (1939). [108] U. Brosa, S. Grossmann, and A. Müller, “Nuclear fission process,” Phys. Rep. 197, 167–262 (1990). [109] J. Randrup and P. Möller, “Brownian shape motion on five-dimensional potential-energy surfaces: Nuclear fission-fragment mass distributions,” Phys. Rev. Lett. 106, 132503 (2011). [110] H. Goutte, J. F. Berger, P. Casoli, and D. Gogny, “Microscopic approach of fission dynamics applied to fragment kinetic energy and mass distributions in 238U,” Phys. Rev. C 71, 024316 (2005). [111] N. Schunck and L. M. Robledo, “Microscopic theory of nuclear fission: a review,” Rep. Prog. Phys. 79 , 116301 (2016). [112] M. A. Nielsen and I. L. Chuang, Quantum Computation and Quantum Information, 10th Anniversary Ed. (Cambridge University Press, 2010). [113] D. Sels and A. Polkovnikov, “Minimizing irreversible losses in quantum systems by local counterdiabatic driving,” Proc. Natl. Acad. Sci. USA 114, E3909–E3916 (2017). [114] M. Demirplak and S. A. Rice, “Adiabatic population transfer with control fields,” J. Phys. Chem. A 107 , 9937–9945 (2003). [115] M. V. Berry, “Transitionless quantum driving,” J. Phys. A 42, 365303 (2009). [116] J. Koch, T. M. Yu, J. Gambetta, A. A. Houck, D. I. Schuster, J. Majer, A. Blais, M. H. Devoret, S. M. Girvin, and R. J. Schoelkopf, “Charge-insensitive qubit design derived from the Cooper pair box,” Phys. Rev. A 76 , 042319 (2007). [117] P. Krantz, M. Kjaergaard, F. Yan, T. P. Orlando, S. Gustavsson, and W. D. Oliver, “A quantum engineer’s guide to superconducting qubits,” Appl. Phys. Rev. 6, 021318 (2019). [118] G. Lüders, “Über die Zustandsänderung durch den Meßprozeß,” Ann. Phys. (Leipzig) 443 , 322–328 (1951). [119] K. Kraus, States, Effects, and Operations: Fundamental Notions of Quantum Theory (Springer, 1983). 67 [120] M.-D. Choi, “Completely positive linear maps on complex matrices,” Linear Algebra Appl. 10 , 285–290 (1975). [121] E. B. Davies and J. T. Lewis, “An operational approach to quantum probability,” Commun. Math. Phys. 17, 239–260 (1970). [122] P. Busch, P. J. Lahti, and P. Mittelstaedt, The Quantum Theory of Measurement, 2nd ed. (Springer, 1996). [123] A. S. Holevo, Probabilistic and Statistical Aspects of Quantum Theory (North–Holland, 1982). [124] M. A. Naimark, “Spectral functions of a symmetric operator,” Izv. Akad. Nauk SSSR Ser. Mat. 4 , 277–318 (1940) [in Russian]. [125] W. F. Stinespring, “Positive functions on C∗-algebras,” Proc. Amer. Math. Soc. 6, 211–216 (1955). [126] A. Peres, Quantum Theory: Concepts and Methods (Kluwer, 1995). [127] R. Alicki and K. Lendi, Quantum Dynamical Semigroups and Applications, 2nd ed. (Springer, 2007). [128] H. Spohn, “Kinetic equations from Hamiltonian dynamics: Markovian limits,” Rev. Mod. Phys. 52 , 569–615 (1980). [129] C. W. Gardiner and P. Zoller, Quantum Noise, 3rd ed. (Springer, 2004). [130] A. J. Leggett, S. Chakravarty, A. T. Dorsey, M. P. A. Fisher, A. Garg, and W. Zwerger, “Dynamics of the dissipative two–state system,” Rev. Mod. Phys. 59, 1–85 (1987). [131] A. O. Caldeira and A. J. Leggett, “Influence of damping on quantum interference: An exactly soluble model,” Physica A 121, 587–616 (1983/85). [132] B. L. Hu, J. P. Paz, and Y. Zhang, “Quantum Brownian motion in a general environment: Exact master equation with nonlocal dissipation and colored noise,” Phys. Rev. D 45, 2843–2861 (1992). [133] C. Cohen–Tannoudji, J. Dupont–Roc, and G. Grynberg, Atom–Photon Interactions: Basic Processes and Applications (Wiley, 1992). [134] M. B. Plenio and P. L. Knight, “The quantum–jump approach to dissipative dynamics in quantum optics,” Rev. Mod. Phys. 70, 101–144 (1998). [135] H. M. Wiseman and L. Diósi, “Complete parameterization and invariance of diffusive quantum trajectories for Markovian open systems,” Chem. Phys. 268, 91–104 (2001). [136] B. Misra and E. C. G. Sudarshan, “The Zeno’s paradox in quantum theory,” J. Math. Phys. 18 , 756–763 (1977). [137] W. M. Itano, D. J. Heinzen, J. J. Bollinger, and D. J. Wineland, “Quantum Zeno effect,” Phys. Rev. A 41 , 2295–2300 (1990). [138] V. P. Belavkin, “Quantum continuous measurements, filtering and control,” in Information Complexity and Control in Quantum Physics, eds. A. Blaquière et al. (Springer, 1987/1992), pp. 311–329. [139] T. W. B. Kibble, “Geometrization of quantum mechanics,” Commun. Math. Phys. 65, 189–201 (1979). [140] M. Nakahara, Geometry, Topology and Physics, 2nd ed. (Taylor & Francis, 2003). [141] A. Edelman, T. A. Arias, and S. T. Smith, “The geometry of algorithms with orthogonality constraints,” SIAM J. Matrix Anal. Appl. 20, 303–353 (1998). [142] B. Simon, “Holonomy, the quantum adiabatic theorem, and Berry’s phase,” Phys. Rev. Lett. 51 , 2167–2170 (1983). [143] A. Shapere and F. Wilczek (eds.), Geometric Phases in Physics (World Scientific, 1989). [144] W. K. Wootters, “Statistical distance and Hilbert space,” Phys. Rev. D 23, 357–362 (1981). [145] L. Campos Venuti and P. Zanardi, “Quantum critical scaling of the geometric tensors,” Phys. Rev. Lett. 99 , 095701 (2007). [146] S.-J. Gu, “Fidelity approach to quantum phase transitions,” Int. J. Mod. Phys. B 24, 4371–4458 (2010). [147] D. Bures, “An extension of Kakutani’s theorem on infinite product measures to the tensor product of semifinite W∗-algebras,” Trans. Amer. Math. Soc. 135, 199–212 (1969). [148] A. Uhlmann, “Parallel transport and ’quantum holonomy’ along density operators,” Rep. Math. Phys. 24 , 229–240 (1986). [149] D. Petz and C. Sudár, “Geometries of quantum states,” J. Math. Phys. 37, 2662–2673 (1996). [150] S. Teufel, Adiabatic Perturbation Theory in Quantum Dynamics (Springer, 2003). [151] D. J. Thouless, M. Kohmoto, M. P. Nightingale, and M. den Nijs, “Quantized Hall conductance in a two-dimensional periodic potential,” Phys. Rev. Lett. 49, 405–408 (1982). [152] J. E. Avron, R. Seiler, and B. Simon, “Homotopy and quantization in condensed matter physics,” Phys. Rev. Lett. 51, 51–53 (1983). [153] F. D. M. Haldane, “Geometrical description of the fractional quantum Hall effect,” Phys. Rev. Lett. 107 , 116801 (2011). [154] N. Marzari and D. Vanderbilt, “Maximally localized Wannier functions for composite energy bands,” Phys. Rev. B 56, 12847–12865 (1997). [155] R. Resta and S. Sorella, “Electron localization in the insulating state,” Phys. Rev. Lett. 82 , 370–373 (1999). [156] D. Diakonov and V. Y. Petrov, “A formula for the Wilson loop,” Phys. Lett. B 224, 131–135 (1989). [157] L. Mandelstam and I. Tamm, “The uncertainty relation between energy and time in non-relativistic quantum mechanics,” J. Phys. (USSR) 9, 249–254 (1945). [158] Y. Aharonov and D. Bohm, “Significance of electromagnetic potentials in the quantum theory,” Phys. Rev. 115, 485–491 (1959). [159] J. Moody, A. Shapere, and F. Wilczek, “Realizations of magnetic–monopole gauge fields: Diatoms and spin precession,” Phys. Rev. Lett. 56, 893–896 (1986). [160] H. Feshbach, “Unified theory of nuclear reactions,” Ann. Phys. 5 , 357–390 (1958); ibid. 19 , 287–313 (1962). 68 [161] M. R. James and J. E. Gough, “Quantum dissipative systems and feedback control design by interconnection,” IEEE Trans. Autom. Control 57, 2050–2063 (2012). [162] F. Shibata, Y. Takahashi, and N. Hashitsume, “A general derivation of the time–convolutionless master equation,” J. Stat. Phys. 17, 171–187 (1977). [163] H. Mori, “Transport, collective motion, and Brownian motion,” Prog. Theor. Phys. 33, 423–455 (1965). [164] R. Azouit, A. Sarlette, and P. Rouchon, “Adiabatic elimination for open quantum systems with effective Lindblad master equations,” arXiv:1506.00973 (2015). [165] F. Verstraete, M. M. Wolf, and J. I. Cirac, “Quantum computation and quantum–state engineering driven by dissipation,” Nature Phys. 5, 633–636 (2009). [166] S. Mac Lane, Categories for the Working Mathematician, 1st ed. (Springer, 1971/1978). [167] A. Joyal and R. Street, “Braided tensor categories,” Adv. Math. 102, 20–78 (1993). [168] J. C. Baez and J. Dolan, “Higher–dimensional algebra and topological quantum field theory,” J. Math. Phys. 36, 6073–6105 (1995/1998). [169] P. Zanardi and M. Rasetti, “Noiseless quantum codes,” Phys. Rev. Lett. 79, 3306–3309 (1997). [170] D. A. Lidar, I. L. Chuang, and K. B. Whaley, “Decoherence–free subspaces for quantum computation,” Phys. Rev. Lett. 81, 2594–2597 (1998). [171] S. Jansen, M. B. Ruskai, and R. Seiler, “Bounds for the adiabatic approximation with applications to quantum computation,” J. Math. Phys. 48, 102111 (2007). [172] T. Albash and D. A. Lidar, “Quantum adiabatic Markovian master equations,” New J. Phys. 14 , 123016 (2012). [173] L. C. Venuti and D. A. Lidar, “Adiabatic theorem for open quantum systems,” Phys. Rev. A 100 , 012112 (2019). [174] G. A. Hagedorn and A. Joye, “Elementary exponential error estimates for the adiabatic approximation,” J. Math. Anal. Appl. 267, 235–246 (2002). [175] J. E. Avron, R. Seiler, and L. G. Yaffe, “Adiabatic theorems and applications to the quantum Hall effect,” Commun. Math. Phys. 110, 33–49 (1987). [176] R. Bellman, Stability Theory of Differential Equations (McGraw–Hill, 1953). [177] D. C. Brody and L. P. Hughston, “Geometric quantum mechanics,” J. Geom. Phys. 38, 19–53 (2001). [178] L. Schlesinger, “Gauge invariance and geometric phase,” J. Math. Phys. 39, 5016–5023 (1968). [179] C. A. Mead and D. G. Truhlar, “On the determination of Born–Oppenheimer nuclear motion wave functions including complications due to conical intersections and identical nuclei,” J. Chem. Phys. 70 , 2284–2296 (1979). [180] E. Sjöqvist, A. K. Pati, A. Ekert, J. S. Anandan, M. Ericsson, D. K. L. Oi, and V. Vedral, “Geometric phases for mixed states in interferometry,” Phys. Rev. Lett. 85, 2845–2849 (2000). [181] J. Dittmann, “The Bures metric on GL(n, C),” Acta Appl. Math. 63, 23–30 (2000). [182] S. Pancharatnam, “Generalized theory of interference and its applications,” Proc. Indian Acad. Sci. A 44 , 247–262 (1956). [183] J. Samuel and R. Bhandari, “General setting for Berry’s phase,” Phys. Rev. Lett. 60, 2339–2342 (1988). [184] P.-A. Absil, R. Mahony, and R. Sepulchre, Optimization Algorithms on Matrix Manifolds (Princeton University Press, 2008). [185] S. Helgason, Differential Geometry, Lie Groups, and Symmetric Spaces (Academic Press, 1978). [186] W. H. Zurek, “Pointer basis of quantum apparatus: Into what mixture does the wave packet collapse?,” Phys. Rev. D 24, 1516–1525 (1981). [187] W. H. Zurek, “Environment-induced superselection rules,” Phys. Rev. D 26, 1862–1880 (1982). [188] T. Gorin, T. Prosen, T. H. Seligman, and M. Žnidarič, “Dynamics of Loschmidt echoes and fidelity decay,” Phys. Rep. 435, 33–156 (2006). [189] R. P. Feynman and F. L. Vernon, “The theory of a general quantum system interacting with a linear dissipative system,” Ann. Phys. 24, 118–173 (1963). [190] U. Weiss, Quantum Dissipative Systems, 2nd ed. (World Scientific, 1999). [191] R. Kubo, “Stochastic Liouville equations,” J. Math. Phys. 4, 174–183 (1963). [192] A. A. Clerk, M. H. Devoret, S. M. Girvin, F. Marquardt, and R. J. Schoelkopf, “Introduction to quantum noise, measurement, and amplification,” Rev. Mod. Phys. 82, 1155–1208 (2010). [193] G. S. Uhrig, “Keeping a quantum bit alive by optimized π -pulse sequences,” Phys. Rev. Lett. 98 , 100504 (2007). [194] L. Cywiński, R. M. Lutchyn, C. P. Nave, and S. Das Sarma, “How to enhance dephasing time in superconducting qubits,” Phys. Rev. B 77, 174509 (2008). [195] G. A. Álvarez and D. Suter, “Measuring the spectrum of colored noise by dynamical decoupling,” Phys. Rev. Lett. 107, 230501 (2011). [196] E. Paladino, Y. M. Galperin, G. Falci, and B. L. Altshuler, “1 /f noise: Implications for solid-state quantum information,” Rev. Mod. Phys. 86, 361–418 (2014). [197] G. Ithier et al., “Decoherence in a superconducting quantum bit circuit,” Phys. Rev. B 72, 134519 (2005). [198] N. Gisin and I. C. Percival, “The quantum-state diffusion model applied to open systems,” J. Phys. A 25 , 5677–5691 (1992). [199] C. W. Gardiner and M. J. Collett, “Input and output in damped quantum systems: Quantum stochastic differential equations and the master equation,” Phys. Rev. A 31, 3761–3774 (1985). [200] R. L. Hudson and K. R. Parthasarathy, “Quantum Itô’s formula and stochastic evolutions,” Commun. Math. 69 Phys. 93, 301–323 (1984). [201] K. Hornberger, “Master equation for a quantum particle in a gas,” Phys. Rev. Lett. 90, 160401 (2003). [202] E. Joos and H. D. Zeh, “The emergence of classical properties through interaction with the environment,” Z. Phys. B 59, 223–243 (1985). [203] H.-P. Breuer, E.-M. Laine, and J. Piilo, “Measure for the degree of non-Markovian behavior of quantum processes in open systems,” Phys. Rev. Lett. 103, 210401 (2009). [204] I. de Vega and D. Alonso, “Dynamics of non-Markovian open quantum systems,” Rev. Mod. Phys. 89 , 015001 (2017). [205] A. Rivas and S. F. Huelga, Open Quantum Systems: An Introduction (Springer, 2011). [206] H. Ollivier, D. Poulin, and W. H. Zurek, “Objective properties from subjective quantum states: Environment as a witness,” Phys. Rev. Lett. 93, 220401 (2004). [207] J. C. Garrison and E. M. Wright, “Complex geometrical phases for dissipative systems,” Phys. Lett. A 128 , 177–181 (1988). [208] D. M. Tong, E. Sjöqvist, L. C. Kwek, and C. H. Oh, “Kinematic approach to the mixed state geometric phase in nonunitary evolution,” Phys. Rev. Lett. 93, 080405 (2004). [209] L. Bouten, R. van Handel, and M. R. James, “An introduction to quantum filtering,” SIAM J. Control Optim. 46, 2199–2241 (2007). [210] K. Jacobs, Quantum Measurement Theory and its Applications (Cambridge University Press, 2014). [211] J. F. Poyatos, J. I. Cirac, and P. Zoller, “Quantum reservoir engineering with laser cooled trapped ions,” Phys. Rev. Lett. 77, 4728–4731 (1996). [212] S. Diehl, A. Micheli, A. Kantian, B. Kraus, H. P. Büchler, and P. Zoller, “Quantum states and phases in driven open quantum systems with cold atoms: Dissipative state preparation,” Nature Phys. 4 , 878–883 (2008). [213] V. Gritsev and A. Polkovnikov, “Dynamical quantum Hall effect in parameter space,” Proc. Natl. Acad. Sci. USA 109, 6457–6462 (2012). [214] R. Capote et al., “Prompt fission neutron spectra: IAEA coordinated research project,” Eur. Phys. J. A 52 , 160 (2016). [215] K.-H. Schmidt and B. Jurado, “Review on the progress in nuclear fission: Experimental methods and theoretical descriptions,” Rep. Prog. Phys. 81, 106301 (2018). [216] A covariant re–expression involves Tr(F12F12)for composite subspaces, but (55) is the minimal statement sufficient for diagnostics in 2D control planes. [217] For unequal edge times TΩ, T∆ one finds Γ A− Γ B = η (∆Ω∆∆) 1 T∆∂Ωg∆∆ −1 TΩ∂∆gΩΩ + O ( k ∆ k3 ). The equal–time choice yields (129). [218] One may include charge polarization and odd–even staggering by additional foldings; these refinements do not affect the conceptual points made here. [219] Microscopic approaches may allow for post–scission emission and Coulomb acceleration; the identity mapping is nevertheless made at the crossing of S. [220] Explicitly, DΩu∆ = ∂Ωu∆ +( ∇u∆·n ) uΩ + extrinsic terms ; only the normal derivative enters the commutator below. [221] In the limit of an abrupt collapse at the terminal scission time, h ( t ) →δ ( t−tsc )and (193) reduces to the standard flux rule, recovering path–independence at fixed endpoints.