Searcharxiv⌕ Search

arXiv subjects

Andrej Novikov

Publications and source records attributed to Andrej Novikov.

9 recordsLinked to original sources

Michelson and Logarithmic Operator Contrasts Spectral Geometry, Tracial Extensions, and Random Products

The Michelson contrast of a positive invertible operator is a bounded coordinate for its condition number. Its logarithmic lift, $\operatorname{OLC}(A)=\frac12\log(\|A\|\|A^{-1}\|)$, connects spectral spread with projective Thompson geometry. This preprint collects the Novikov--Nikolaeva series and the associated single-author extensions by Novikov into one treatment. We develop the equality criterion for product subadditivity, its dimension and Jordan-rank thresholds, quantum-state examples, and contrasts associated with unitarily invariant norms. The latter have a sharp additive defect, while the full Schatten family records a logarithmic spectral profile up to homometry. Tracial extensions give characterisations of traces, dimension-sensitive defects, and spectral-edge asymptotics. Unbounded operators are described through critical exponents and finite sections. Tensor, matrix and free independence lead to distinct growth laws for products. Finally, symmetric cones provide a setting for the equality criterion, Lorentz boosts, and Jordan products. Classical ingredients are distinguished from the refinements developed in the series; conjectural asymptotic statements remain explicitly labelled.

math.OA↗

How characterisations of traces and central elements fail: sharp violation constants via michelson contrast and noncentrality

A positive functional is a trace if and only if it satisfies any one of many inequalities, the first being Gardner's $|φ(X)|\leφ(|X|)$; positive central elements are characterised similarly. We ask how strongly these characterisations fail. For $φ=\mathrm{Tr}(ρ\,\cdot)$ on a matrix algebra the natural scale is the Michelson contrast $δ$ of the density. For Gardner's inequality the sharp constant depends on $δ$ and on the contrast $δ'$ of $|X|$ only through their product: it is $(1-δ^2δ'^2)^{-1/2}$, the secant of the angle whose sine is $δδ'$. On a finite von Neumann algebra the contrast is the wrong parameter, since every central density gives a trace; the governing parameter is the noncentrality, defined through the central condition number $κ_Z(ρ)$ of the central envelopes of $ρ$ (the largest and smallest elements of Halpern's central spectrum). A transfer principle lifts sharp matrix constants to exact constants in terms of $κ_Z$ in every finite von Neumann algebra. For 38 trace characterisations the constant is exact in 23 cases, has proved bounds in 13 and is known only numerically in two; for instance it is $\cosh(\frac12\lnκ_Z)$ for Gardner's inequality and for the subadditivity of the modulus. On matrix algebras the finite exact violations vanish at the trace to the orders $δ$, $δ^2$ or $δ^4$, and outside factors and abelian algebras no function of the contrast describes the constants. We also find the sharp constants of the Novikov-Tikhonov characterisations of central elements, sharp bounds for three unnormalised trace functionals, and sharp contrast thresholds below which subadditivity of the square root and the power-mean inequality cannot fail. A numerical search, certified as a lower bound for each supremum, supports conjectured values for the remaining inequalities.

math.OA↗

Finite-Dimensional Type I von Neumann Algebras in PyTorch: A GPU-Accelerated Framework for Random Block-Diagonal Operators

torch_vn_algebra is a PyTorch library for numerical work with finite-dimensional von Neumann algebras, i.e. direct sums of matrix algebras. These algebras are central to operator algebra theory and, via commutants of symmetries, describe quantum systems with conserved charges or superselection rules. To our knowledge, no numerical software treats them as a data type with their centre, family of traces, conditional expectations, inclusions and functional calculus. The library fills the gap between computer algebra (exact but small-scale) and quantum simulators (large-scale but limited to Hilbert-space matrices): it handles batches of operators on CPUs and GPUs with automatic differentiation, and converts operators, states and channels. Key features: Compact tensor representation: 4D tensor (B, C, kmax, kmax), where B is batch size (e.g. Monte Carlo samples), C is the number of summands, and kmax is the max block size, enabling batched GPU processing. Lazy evaluation: matrix data materialized only on access, avoiding memory waste. Random operator generation: custom eigenvalue distributions; Haar, SU(n), COE, CSE unitary ensembles. Functional calculus and spectral methods: matrix absolute value, square root, inverse, entropy via SVD; automatic switch between exact diagonalization (block size up to 256) and power iteration for extreme eigenvalues. Three trace functionals: Tr_blunt (sum of block traces), Tr_norm (dimension-normalized), tau_vN (faithful normal tracial state). Performance: on a Tesla P100 GPU, processing 20,000 samples of 100x100 operators yields up to 30x speedup over traditional CPU workflows. Availability: open-source at https://gitlab.com/a.hobukob/von_neumann_type_i/

cs.MS↗

Separating Statistical Inference from Experimental Runtime: An Architecture for Sequential Scientific Analysis

Sequential statistical procedures are increasingly embedded in scientific systems in which observations arrive incrementally from heterogeneous sources. In such systems the statistical update is only one part of the computation: observations must also be acquired, aligned across inputs, ordered, delivered, persisted, and replayed, and results must be passed to systems that display them or act on them. Sequential methods are nevertheless missing from the general software toolkit. They run inside proprietary experimentation platforms and commercial statistical packages, or exist as offline functions over complete data, and the batch fit interface that dominates machine-learning libraries has no place for a procedure that listens, keeps state, and decides when to stop. We propose an architecture that separates statistical inference from data topology, execution runtime, run lifecycle, persistence, and external action. Its central abstraction is the Sequential Procedure: a state-transition function that maps an inferential state, a valid statistical input, and a statistical random state to a new state and a scientific output. The procedure does not define how observations are acquired, paired, scheduled, stored, or acted upon. We specify the architecture semantically rather than through an object model, derive seven conformance invariants that any implementation can be tested against (among them synchronous/asynchronous, checkpoint, replay, and batch equivalence), and describe a dependency-free reference implementation in which all seven hold bitwise. Lorden's 2-SPRT for the Kiefer-Weiss problem serves as the reference procedure; the same contract accommodates sequential estimators, Bayesian filters, change-point detectors, and stateful predictive models.

math.ST↗

Upper and lower limits of scales and the interchange of quantifiers

We define lower and upper limits for a family of structured subobjects of a fixed ambient object, as the direct limit of the meets over tails and the inverse limit of the joins over tails, and construct the canonical comparison morphism between them. The family need not be a diagram: its members are related only through the ambient object. We distinguish two senses in which the comparison may be invertible, before and after forgetting the structure, and prove that they coincide exactly when the forgetful functor reflects isomorphisms; they therefore agree for sets, for vector spaces and for Banach spaces, and separate for locally convex spaces. For countable scales of Banach spaces the separation cannot occur when the upper limit is ultrabornological, so under those hypotheses the obstruction is localised in the topology of that limit. Intermediate conditions correspond to factorisations of the forgetful functor, of which bornological convergence is one. In the cases that matter the comparison is an interchange of quantifiers: membership in the lower limit means that one value of the second index serves every value of the first, and membership in the upper limit that every value of the first is served by some value of the second. We exhibit this for spaces of test functions, for weighted Sobolev spaces, for Laurent series and for the adeles, and discuss three further conditions that suggest themselves.

math.OA↗

Operator Michelson Contrast: Logarithmic Parametrisation, Subadditivity, and a Sharp Noncommutativity Threshold

We study the operator Michelson contrast Delta(A) = (kappa(A)-1)/(kappa(A)+1), with kappa(A)=||A|| ||A^{-1}||, for positive invertible operators, and introduce its logarithmic parametrization via arctanh(Delta(A)) = (1/2) ln kappa(A). This defines the Operator Logarithmic Contrast (OLC), extended to all invertible operators via polar decomposition. We prove OLC is subadditive under multiplication: OLC(AB) <= OLC(A)+OLC(B). For positive invertible A, we identify OLC(A) exactly as the projective Thompson distance from the identity ray to the ray of A. We establish a sharp dimensional threshold for equality in subadditivity: for dimensions 2 and 3, equality forces commutativity, while n=4 is the smallest dimension admitting noncommuting equality examples. Moreover, equality still forces commutativity in any total dimension if the operators factor as pure Kronecker products X = tensor X_i and Y = tensor Y_i with each factor dimension <= 3. We derive contraction inequalities, sum bounds, trace estimates for density operators, Lipschitz continuity (including singular limits), and limiting behavior for Cesaro means and infinite products. For density operators, we obtain an exact closed-form bijection between OLC and von Neumann entropy for qubits; prove this bijection does not exist for d >= 3; and establish upper/lower entropy envelopes at fixed condition number, showing the admissible range [S_min(kappa), S_max(kappa)] satisfies ln 2 <= S_max(kappa) < ln 3 for every kappa > 1. The qubit case recovers, as an operator lift, the classical Michelson-Jensen-Shannon equivalence of Bruni, Rossi, and Vitulano, while a counterexample shows this equivalence fails for d > 2 as a direct consequence of the non-bijectivity.

math.QA↗

$L_1$-space for a positive operator affiliated with von Neumann algebra

In this paper we suggest an approach for constructing an L1-type space for a positive selfadjoint operator affiliated with von Neumann algebra. For such operator we intro- duce a seminorm, and prove that it is a norm if and only if the operator is injective. For this norm we construct an L1 -type space as the complition of the space of hermitian ultraweakly continuous linear functionals on von Neumann algebra, and represent L1- type space as a space of continuous linear functionals on the space of special sesquilinear forms. Also, we prove that L1 -type space is isometrically isomorphic to the predual of von Neumann algebra in a natural way. We give a small list of alternate definitions of the seminorm, and a special definition for the case of semifinite von Neumann algebra, in particular. We study order properties of L1-type space, and demonstrate the con- nection between semifinite normal weights and positive elements of this space. At last, we construct a similar L-space for the positive element of C*-algebra, and study the connection between this L-space and the L1 -type space in case when this C*-algebra is a von Neumann algebra.

math.OA↗