Continuous degrees of freedom The quantum systems we have discussed so far have been described by finite-dimensional Hilbert spaces. Basic measurements on such systems have a finite number of possible outcomes, and a quantum state predicts a discrete probability distribution over these. Now we wish to extend our theory to handle systems with one or more continuous degrees of freedom, such as the position of a particle that can move in one dimension. This will require an extension of our theory to Hilbert spaces of infinite dimension, and to systems with continuous observables. There is a philosophical issue here. How do we know that there really are infinitely many distinct locations for a particle? The short answer is, we don't. It might be that space itself is both discrete (at the tiniest scales) and bounded (at the largest), so that the number of possible locations of a particle is some very large but finite number. If this is the case, then the continuum model for space is nothing more than a convenient approximation. Infinity is just a simplified way of describing a quantity that is immense, but still finite. In this section, we will adopt this view of infinity. We will imagine that any continuous variable is really an approximation of a “true” discrete variable. This idea will motivate the continuous quantities and operations that we need.
This paper argues that the quantum measurement problem is not a single explanatory gap but a collection of distinct tasks. We separate detector representability, probability structure, single-event actuality, and outcome-conditioned refinement, and distinguish all four from causal registration, the physical criterion by which an already-instantiated event becomes a completed measurement record. These are explanatory roles within one measurement phenomenon rather than successive temporal stages. The analysis begins from the contrast between a global quantum description containing several detector-correlated components and the single entry that appears in a realized measurement history. Rather than compressing this contrast into an unexplained transition often labeled “collapse,” we ask which specific explanatory constraints must be satisfied before the relation between the two descriptions can be diagnosed precisely. A finite-resolution alternating position–momentum history is used throughout to separate questions of record formation, probability, actuality, causal registration, and outcome-conditioned refinement. A continuum analysis shows that making the registered history arbitrarily dense does not by itself recover a classical trajectory. Several parts of the architecture are exact once their inputs are specified. A detector representation induces an exact quotient; a specified acquisition channel and record family induce system-side effects; and a specified quantum instrument determines a family of subnormalized conditional branches before any one outcome is instantiated. Given a realized outcome label, the same instrument determines the corresponding normalized outcome-conditioned state. A finite-dimensional acquisition model explicitly realizes the representability chain, while a record-algebra indistinguishability result shows that removing coherence between detector-record sectors can leave all expectation values in the lifted record algebra unchanged even when that coherence remains detectable in a richer context. These results sharpen, but do not dissolve, the measurement problem. Record structure does not derive Born weights; a probability law does not itself constitute the event realized in one run; a mathematically available conditional branch family does not select one branch condition as actual; and causal onset, stabilization, registration, and conditional refinement all presuppose or follow that event rather than produce it. The resulting constraint audit tests—without assuming—the hypothesis that global quantum multiplicity and detector-record definiteness may belong to different representational levels. The main conclusion is not a new collapse dynamics but a localization of the remaining explanatory burden. Once representability, probability, causal registration, and conditional refinement are assigned their proper roles, the unresolved actuality transition can no longer be hidden inside the single label “collapse.” In the typed measurement architecture considered here, no pre-event operation has the mathematical role of producing an individual event token without additional event-producing structure. This is version 1.0 of the manuscript.
This is a mathematics paper. It constructs a syndrome algebra for differentiable parametric systems by generalising the syndrome-measurement principle used in quantum error correction (QEC) from a discrete binary setting to a continuous real-valued one. The result is a tool for diagnosing the response of a differentiable system to weight-space perturbations; its concrete application to neural network reliability is the subject of the companion paper "Three Measurable Failure Modes of Large Language Models", this paper is focused on its mathematical formalism. The starting point is the
[[5,1,3]] perfect quantum error-correcting code, whose parity-check matrix maps single-qubit errors to four-bit syndromes. Three replacements take this construction to the continuous setting. (1) Field replacement: the finite-field arithmetic over
F2 is replaced by row-wise
ℓ2-normalisation over
R. (2) Basis replacement: the symplectic Pauli error basis is replaced by the right singular vectors of the weight matrix (the SVD basis). (3) Map replacement: the parity-check matrix is replaced by the averaged Jacobian of the logit map at probe inputs. Under these three substitutions, the syndrome table becomes
S=N(Jˉ⋅V)⊤, and the discrete
[[5,1,3]] code is recovered as the
q=2 special case. Seven theorems establish the algebra's main properties: Theorem 1: the Jacobian is the syndrome table in the linear regime. Theorem 2: oracle correction is exact on the linear path. Theorem 3: crossing error (correction in a wrong direction) is positive in expectation. Theorem 4: the null space of the weight matrix is the confabulation subspace. Theorem 5: multi-layer measurement achieves identifiability. Theorem 6 (Jacobian Uncertainty Principle): the product of dictionary variance and observation variance is bounded below by a quantity proportional to the per-direction Jacobian variance. Theorem 7 (Reliability Principle): for any
[[n,k,d]]q code, normalised robustness, capacity, and expressiveness satisfy a structural trade-off that no architecture can escape. Additional notes: This paper presents the mathematical foundation for the syndrome algebra framework. Application and experimental validation on neural networks is provided in the companion paper Three Measurable Failure Modes of Large Language Models (see related identifiers). The paper is self-contained: the complete derivation from the
[[5,1,3]] quantum error-correcting code to the continuous real-valued syndrome table is provided, with all proofs in the main body; no prior reading of the experimental companion is required.