-
Hamiltonian Reordering for Shallower Trotterization Circuits
Authors:
Cédric Ho Thanh
Abstract:
Quantum simulation is a popular application of quantum computing, but its practical realization is hindered by the technical limitations of current devices. In this work, we focus on preprocessing Hamiltonians before Trotterization to generate shallower evolution circuits, which are less prone to noise and decoherence. Specifically, we apply graph coloring techniques to reorder Pauli terms and inc…
▽ More
Quantum simulation is a popular application of quantum computing, but its practical realization is hindered by the technical limitations of current devices. In this work, we focus on preprocessing Hamiltonians before Trotterization to generate shallower evolution circuits, which are less prone to noise and decoherence. Specifically, we apply graph coloring techniques to reorder Pauli terms and increase "gate parallelism". We benchmark two coloring algorithms, and report the depth reduction and computational overhead. Then, we examine how these optimized circuits affect the performance of the Quantum Approximate Optimization Algorithm (QAOA). Our results show that shallower circuits lead to faster convergence and reach higher energy levels compared to their non-reordered counterparts.
△ Less
Submitted 14 March, 2025;
originally announced March 2025.
-
Improving Fine-Tuning with Latent Cluster Correction
Authors:
Cédric Ho Thanh
Abstract:
The existence of salient semantic clusters in the latent spaces of a neural network during training strongly correlates its final accuracy on classification tasks. This paper proposes a novel fine-tuning method that boosts performance by optimising the formation of these latent clusters, using the Louvain community detection algorithm and a specifically designed clustering loss function. We presen…
▽ More
The existence of salient semantic clusters in the latent spaces of a neural network during training strongly correlates its final accuracy on classification tasks. This paper proposes a novel fine-tuning method that boosts performance by optimising the formation of these latent clusters, using the Louvain community detection algorithm and a specifically designed clustering loss function. We present preliminary results that demonstrate the viability of this process on classical neural network architectures during fine-tuning on the CIFAR-100 dataset.
△ Less
Submitted 21 January, 2025;
originally announced January 2025.
-
Opetopic algebras III: Presheaf models of homotopy-coherent opetopic algebras
Authors:
Cédric Ho Thanh,
Chaitanya Leena Subramaniam
Abstract:
For $Λ$ the category of free opetopic algebras, we construct a model structure \emph{à la Cisinski} on the category of presheaves over $Λ$, and show that is is equivalent to opetopic complete Segal spaces. This generalizes the results of Joyal and Tierney in the simplicial case, and of Cisinski and Moerdijk in the planar dendroidal case.
For $Λ$ the category of free opetopic algebras, we construct a model structure \emph{à la Cisinski} on the category of presheaves over $Λ$, and show that is is equivalent to opetopic complete Segal spaces. This generalizes the results of Joyal and Tierney in the simplicial case, and of Cisinski and Moerdijk in the planar dendroidal case.
△ Less
Submitted 22 January, 2020;
originally announced January 2020.
-
Opetopic algebras I: Algebraic structures on opetopic sets
Authors:
Cédric Ho Thanh,
Chaitanya Leena Subramaniam
Abstract:
We define a family of structures called "opetopic algebras", which are algebraic structures with an underlying opetopic set. Examples of such are categories, planar operads, and Loday's combinads over planar trees. Opetopic algebras can be defined in two ways, either as the algebras of a "free pasting diagram" parametric right adjoint monad, or as models of a small projective sketch over the categ…
▽ More
We define a family of structures called "opetopic algebras", which are algebraic structures with an underlying opetopic set. Examples of such are categories, planar operads, and Loday's combinads over planar trees. Opetopic algebras can be defined in two ways, either as the algebras of a "free pasting diagram" parametric right adjoint monad, or as models of a small projective sketch over the category of opetopes. We define an opetopic nerve functor that fully embeds each category of opetopic algebras into the category of opetopic sets. In particular, we obtain fully faithful opetopic nerve functors for categories and for planar coloured Set-operads.
This paper is the first in a series aimed at using opetopic spaces as models for higher algebraic structures.
△ Less
Submitted 22 January, 2020; v1 submitted 3 November, 2019;
originally announced November 2019.
-
Syntactic approaches to opetopes
Authors:
Pierre-Louis Curien,
Cédric Ho Thanh,
Samuel Mimram
Abstract:
Opetopes are algebraic descriptions of shapes corresponding to compositions in higher dimensions. As such, they offer an approach to higher-dimensional algebraic structures, and in particular, to the definition of weak $ω$-categories, which was the original motivation for their introduction by Baez and Dolan. They are classically defined inductively (as free operads in Leinster's approach, or as z…
▽ More
Opetopes are algebraic descriptions of shapes corresponding to compositions in higher dimensions. As such, they offer an approach to higher-dimensional algebraic structures, and in particular, to the definition of weak $ω$-categories, which was the original motivation for their introduction by Baez and Dolan. They are classically defined inductively (as free operads in Leinster's approach, or as zoom complexes in the formalism of Kock et al.), using abstract constructions making them difficult to manipulate with a computer. In this paper, we present two purely syntactic descriptions of opetopes as sequent calculi, the first using variables to implement the compositional nature of opetopes, the second using a calculus of higher addresses. We prove that well-typed sequents in both systems are in bijection with opetopes as defined in the more traditional approaches. Additionally, we propose three variants to describe opetopic sets. We expect that the resulting structures can serve as natural foundations for mechanized tools based on opetopes.
△ Less
Submitted 14 March, 2019;
originally announced March 2019.
-
The equivalence between many-to-one polygraphs and opetopic sets
Authors:
Cédric Ho Thanh
Abstract:
From the polynomial approach to the definition of opetopes of Kock et al., we derive a category of opetopes, and show that its set-valued presheaves, or opetopic sets, are equivalent to many-to-one polygraphs. As an immediate corollary, we establish that opetopic sets are equivalent to multitopic sets, introduced and studied by Harnick et al, and we also address an open question of Henry.
From the polynomial approach to the definition of opetopes of Kock et al., we derive a category of opetopes, and show that its set-valued presheaves, or opetopic sets, are equivalent to many-to-one polygraphs. As an immediate corollary, we establish that opetopic sets are equivalent to multitopic sets, introduced and studied by Harnick et al, and we also address an open question of Henry.
△ Less
Submitted 21 September, 2021; v1 submitted 22 June, 2018;
originally announced June 2018.