Search arXivSearch

arXiv · 2212.06615

Category Theory for Quantum Natural Language Processing

Abstract

This thesis introduces quantum natural language processing (QNLP) models based on a simple yet powerful analogy between computational linguistics and quantum mechanics: grammar as entanglement. The grammatical structure of text and sentences connects the meaning of words in the same way that entanglement structure connects the states of quantum systems. Category theory allows to make this language-to-qubit analogy formal: it is a monoidal functor from grammar to vector spaces. We turn this abstract analogy into a concrete algorithm that translates the grammatical structure onto the architecture of parameterised quantum circuits. We then use a hybrid classical-quantum algorithm to train the model so that evaluating the circuits computes the meaning of sentences in data-driven tasks. The implementation of QNLP models motivated the development of DisCoPy (Distributional Compositional Python), the toolkit for applied category theory of which the first chapter gives a comprehensive overview. String diagrams are the core data structure of DisCoPy, they allow to reason about computation at a high level of abstraction. We show how they can encode both grammatical structures and quantum circuits, but also logical formulae, neural networks or arbitrary Python code. Monoidal functors allow to translate these abstract diagrams into concrete computation, interfacing with optimised task-specific libraries. The second chapter uses DisCopy to implement QNLP models as parameterised functors from grammar to quantum circuits. It gives a first proof-of-concept for the more general concept of functorial learning: generalising machine learning from functions to functors by learning from diagram-like data. In order to learn optimal functor parameters via gradient descent, we introduce the notion of diagrammatic differentiation: a graphical calculus for computing the gradients of parameterised diagrams.

Explore related subjects

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Alexis Toumi. 2022-12-13. Category Theory for Quantum Natural Language Processing. https://arxiv.org/abs/2212.06615

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

A Categorical Generalization of Counterpoint

We extend Mazzola's counterpoint model using category theory, generalizing from the category $\mathbf{Set}$ to an arbitrary topos other than $\mathbf{Set}$. This generalization suggests that counterpoint's essential structure depends on specific categorical conditions rather than classical set-theoretic reasoning. A key contribution is identifying sufficient requirements for a well-behaved counterpoint theory in a topos: some version of Zorn's Lemma (GJZL), and two-valuedness and split supports (NS). Within a topos, we introduce (weak) quasidichotomies alongside the classical notion of dichotomy. These structures capture varying degrees of oppositional structure between consonance and dissonance, with weak quasidichotomies preserving the non-Boolean flexibility essential to musical practice while quasidichotomies represent maximal opposition short of complete partition. We prove a generalized counterpoint theorem giving sufficient conditions for the existence of admitted successors. When the ambient topos turns non-zero successor objects into points, admitted succession can be iterated to form counterpoint paths, which may terminate at consonances with no admitted successor. The framework naturally accommodates counterpoint with sets instead of pure pitches, relaxing the ``yes/no'' character of classical consonance definitions and emphasizing context-dependence. Mazzola's model allows a Kuratowski closure operator induced by a polarity, which defines an internal topology enabling algebraic-topological analysis of counterpoint structure. We conclude by showing this construction generalizes to involutive morphisms. This categorical approach provides foundations for understanding both the historical evolution of contrapuntal practice and cross-cultural divergences in interval organization.

math.CT

From 3-crossed modules to Gray-type 4-categories

In this paper, we investigate the relation between the category of 3-crossed modules and the category of Gray-type 4-groups. The notion of a 3-crossed module was first introduced by Arvasi \textit{et al.}, motivated by the question of what kind of algebraic structure completely encodes a homotopy 4-type. On the other hand, from the point of view that higher groups are equivalent to algebraic realizations of higher categories -- as exemplified by the relationship between 2-crossed modules and Gray 3-groups established by Sarikaya--Ulualan -- it had not been clear how the 3-crossed modules of Arvasi \textit{et al.} relate to any higher category. In our previous paper, we proposed a new definition of a 3-crossed module and observed that it admits a natural interpretation in terms of higher categories. In this paper, we make this interpretation precise: we introduce a 4-category, which reduces to a semistrict braided monoidal 2-category when restricted to a single object and a single 1-morphism, and prove that the category of our 3-crossed modules is equivalent to the category of Gray 4-groups, defined as single-object versions of this 4-category in which all morphisms are invertible. We therefore expect that these structures can correctly capture the topological nature of surface knots and higher-dimensional manifolds.

math.CT

Observations on the variety of equationally linear Heyting semilattices

In previous work, we analysed a number of categorical properties, of interest in the context of Janelidze-Márki-Tholen semi-abelian categories, for the variety $\mathsf{HSLat}$ of Heyting semilattices. In this paper, we focus on the subvariety $\mathsf{ELHSLat}$ of equationally linear Heyting semilattices. Our main objective is to show that, unlike $\mathsf{HSLat}$, this category is algebraically coherent. We furthermore prove that it is neither locally algebraically cartesian closed nor cosmash associative.

math.CT