arXiv subjects

Gemma De les Coves

Publications and source records attributed to Gemma De les Coves.

10 recordsLinked to original sources

arXiv2026

The Structure of Emulations in Classical Spin Models: Modularity and Universality

The theory of spin models intersects with condensed matter physics, complex systems, graph theory, combinatorial optimization, computational complexity and neural networks. Many ensuing applications rely on the fact that complicated spin models can be transformed to simpler ones. What is the structure of such transformations? Here, we provide a framework to study and construct emulations between spin models. A spin model is a set of spin systems, and emulations are efficiently computable simulations with arbitrary energy cut-off, where a source spin system simulates a target system if, below the cut-off, the target Hamiltonian is encoded in the source Hamiltonian. We prove that emulations preserve important properties, as they induce reductions between computational problems such as computing ground states, approximating partition functions and approximate sampling from Boltzmann distributions. Emulations are modular (they can be added, scaled and composed), and allow for universality, i.e. certain spin models have maximal reach. We prove that a spin model is universal if and only if it is scalable, closed and functional complete. Because the characterization is constructive, it provides a step-by-step guide to construct emulations. We prove that the 2d Ising model with fields is universal, for which we also provide two new crossing gadgets. Finally, we show that simulations can be computed by linear programs. While some ideas of this work are contained in [1], we provide new definitions and theorems. This framework provides a toolbox for applications involving emulations of spin models.

math-ph↗

arXiv2026

The View from Within: What Can Embedded Observers (Not) Learn?

Physics is usually done from a third-person perspective, as if the world were described from outside. Yet, observers are themselves physical systems within the world they observe. Here, we investigate this tension using a toy model of classical particles. Observers are physical systems characterised by a choice of (i) a manifest variable, whose value constitutes their empirical record, and (ii) ready states, which provide initial information. We ask what such observers (subjects) can learn about the world through interactions that establish a correlation with another system (the object). For each combination of the three types of learning (about the past, the future, or both), manifest variables, and ready states, we determine what the subject can learn. This shows that many subjects face an epistemic horizon---a limitation to what they can learn about the world---even though the model is classical and deterministic. For example, a subject with a complete manifest variable---one whose ontic state is the empirical record---and partial initial information can learn at most half the object's variables, recovering Spekkens' knowledge-balance principle, i.e. an analogue of Heisenberg's uncertainty principle. More generally, we find that subjects face more severe epistemic horizons when predicting than when retrodicting, and that learning by repeatable measurements can be more limited than either prediction or retrodiction alone. Our work provides language and tools to study how the first-person perspective can differ from the third-person perspective beyond the toy model studied here, and invites explanations of the inherent uncertainty in quantum theory from the standpoint of embedded observers.

quant-ph↗

arXiv2026

Universal Spin Models are Universal Approximators in Machine Learning

One of the theoretical pillars that sustain certain machine learning models are universal approximation theorems, which prove that they can approximate all functions from a function class to arbitrary precision. Independently, classical spin models are termed universal if they can reproduce the behavior of any other spin model in their low energy sector. Universal spin models have been characterized via sufficient and necessary conditions, showing that simple models such as the 2d Ising with fields are universal. In this work, we prove that universal spin models are universal approximators of probability distributions. This enables us to leverage the characterization of the former to reveal conditions which are sufficient for universal approximation. Deriving universal approximation theorems thus amounts to verifying these conditions, yielding a unified recipe for universal approximation theorems applicable to a wide range of models. We explicitly test this recipe for restricted and deep Boltzmann machines, as well as for deep belief networks. This work illustrates that independently discovered universality statements may be intimately related, enabling the transfer of results.

cond-mat.dis-nn↗

arXiv2025

The grammar of the Ising model: A new complexity hierarchy

How complex is an Ising model? Usually, this is measured by the computational complexity of its ground state energy problem. Yet, this complexity measure only distinguishes between planar and non-planar interaction graphs, and thus fails to capture properties such as the average node degree, the number of long range interactions, or the dimensionality of the lattice. Herein, we introduce a new complexity measure for Ising models and thoroughly classify Ising models with respect to it. Specifically, given an Ising model we consider the decision problem corresponding to the function graph of its Hamiltonian, and classify this problem in the Chomsky hierarchy. We prove that the language of this decision problem is (i) regular if and only if the Ising model is finite, (ii) constructive context free if and only if the Ising model is linear and its edge language is regular, (iii) constructive context sensitive if and only if the edge language of the Ising model is context sensitive, and (iv) decidable if and only if the edge language of the Ising model is decidable. We apply this theorem to show that the 1d Ising model, the Ising model on generalised ladder graphs, and the Ising model on layerwise complete graphs are constructive context free, while the 2d Ising model, the all-to-all Ising model, and the Ising model on perfect binary trees are constructive context sensitive. This work is a first step in the characterisation of physical interactions in terms of grammars.

cond-mat.stat-mech↗

arXiv2025

Positive Moments Forever: Undecidable and Decidable Cases

We investigate the generalized moment membership problem for matrices, a formulation equivalent to Skolem's problem for linear recurrence sequences. We show decidability for orthogonal, unitary, and real eigenvalue matrices, and undecidability for matrices over certain commutative and non-commutative polynomial rings. As consequences, we deduce that positivity is decidable for simple unitary linear recurrence sequences and undecidable for linear recurrence sequences over commutative polynomial rings. As a byproduct, we also prove a free version of Polya's theorem.

math.AG↗

arXiv2025

Border Ranks of Positive and Invariant Tensor Decompositions: Applications to Correlations

The matrix rank and its positive versions are robust for small approximations, i.e. they do not decrease under small perturbations. In contrast, the multipartite tensor rank can collapse for arbitrarily small errors, i.e. there may be a gap between rank and border rank, leading to instabilities in the optimization over sets with fixed tensor rank. Can multipartite positive ranks also collapse for small perturbations? In this work, we prove that multipartite positive and invariant tensor decompositions exhibit gaps between rank and border rank, including tensor rank purifications and cyclic separable decompositions. We also prove a correspondence between positive decompositions and membership in certain sets of multipartite probability distributions, and leverage the gaps between rank and border rank to prove that these correlation sets are not closed. It follows that testing membership of probability distributions arising from resources like translational invariant Matrix Product States is impossible in finite time. Overall, this work sheds light on the instability of ranks and the unique behavior of bipartite systems.

quant-ph↗

arXiv2024

A Framework for Universality in Physics, Computer Science, and Beyond

Turing machines and spin models share a notion of universality according to which some simulate all others. Is there a theory of universality that captures this notion? We set up a categorical framework for universality which includes as instances universal Turing machines, universal spin models, NP completeness, top of a preorder, denseness of a subset, and more. By identifying necessary conditions for universality, we show that universal spin models cannot be finite. We also characterize when universality can be distinguished from a trivial one and use it to show that universal Turing machines are non-trivial in this sense. Our framework allows not only to compare universalities within each instance, but also instances themselves. We leverage a Fixed Point Theorem inspired by a result of Lawvere to establish that universality and negation give rise to unreachability (such as uncomputability). As such, this work sets the basis for a unified approach to universality and invites the study of further examples within the framework.

cs.CC↗

arXiv2024

Epistemic Horizons From Deterministic Laws: Lessons From a Nomic Toy Theory

Quantum theory has an epistemic horizon, i.e. exact values cannot be assigned simultaneously to incompatible physical quantities. As shown by Spekkens' toy theory, positing an epistemic horizon akin to Heisenberg's uncertainty principle in a classical mechanical setting also leads to a plethora of quantum phenomena. We introduce a deterministic theory - nomic toy theory - in which information gathering agents are explicitly modelled as physical systems. Our main result shows the presence of an epistemic horizon for such agents. They can only simultaneously learn the values of observables whose Poisson bracket vanishes. Therefore, nomic toy theory has incompatible measurements and the complete state of a physical system cannot be known. The best description of a system by an agent is via an epistemic state of Spekkens' toy theory. Our result reconciles us to measurement uncertainty as an aspect of the inseparability of subjects and objects. Significantly, the claims follow even though nomic toy theory is essentially classical. This work invites further investigations of epistemic horizons, such as the one of (full) quantum theory.

quant-ph↗

arXiv2024

An Invitation to Universality in Physics, Computer Science, and Beyond

A universal Turing machine is a powerful concept - a single device can compute any function that is computable. A universal spin model, similarly, is a class of physical systems whose low energy behavior simulates that of any spin system. Our categorical framework for universality (arXiv:2307.06851) captures these and other examples of universality as instances. In this article, we present an accessible account thereof with a focus on its basic ingredients and ways to use it. Specifically, we show how to identify necessary conditions for universality, compare types of universality within each instance, and establish that universality and negation give rise to unreachability (such as uncomputability).

cs.CC↗

arXiv2023

Beyond Operator Systems

Operator systems connect operator algebra, free semialgebraic geometry and quantum information theory. In this work we generalize operator systems and many of their theorems. While positive semidefinite matrices form the underlying structure of operator systems, our work shows that these can be promoted to far more general structures. For instance, we prove a general extension theorem which unifies the well-known homomorphism theorem, Riesz' extension theorem, Farkas' lemma and Arveson's extension theorem. On the other hand, the same theorem gives rise to new vector-valued extension theorems, even for invariant maps, when applied to other underlying structures. We also prove generalized versions of the Choi-Kraus representation, Choi-Effros theorem, duality of operator systems, factorizations of completely positive maps, and more, leading to new results even for operator systems themselves. In addition, our proofs are shorter and simpler, revealing the interplay between cones and tensor products, captured elegantly in terms of star autonomous categories. This perspective gives rise to new connections between group representations, mapping cones and topological quantum field theory, as they correspond to different instances of our framework and are thus siblings of operator systems.

math.OA↗