Searcharxiv⌕ Search

arXiv subjects

Daniel Barter

Publications and source records attributed to Daniel Barter.

12 recordsLinked to original sources

Neural Network Path Optimization for Finding Transition States on a Machine Learned Potential

Transition states govern chemical reaction rates and outcomes, yet locating them remains a central challenge in computational chemistry. Double-ended search methods find transition states given reactant and product structures, but generally rely on discretized path representations developed for density functional theory, where potential energy surface evaluations are expensive. Machine learned interatomic potentials (MLIPs) dramatically reduce this cost and leverage automatic differentiation to access higher-order derivatives efficiently. We introduce Popcornn, a framework to find transition states by optimizing continuous neural network paths atop an MLIP with a parallelized adaptive numerical integrator. This combination eliminates path discretization artifacts, reduces hyperparameter sensitivity, and enables non-local optimization to overcome local features using second-order potential-energy-surface information. Popcornn consistently recovers more correct transition states than standard methods such as nudged elastic band and growing string, across MLIPs on thousands of organic reactions, establishing it as a robust, automation-ready method for high-throughput transition state search. Popcornn's infrastructure is flexible in path objective, integration strategy, and optimization algorithm, making it a practical tool and platform for investigating open questions in MLIP-based path optimization.

physics.chem-ph↗

Machine-learning decoders for fusion-based quantum computing in real time

Fault-tolerant quantum computation requires decoders that combine accuracy, throughput, low latency, and scalability to large numbers of physical qubits. In photonic fusion-based architectures, erasures from photon loss and probabilistic fusion measurements are a dominant error mechanism, motivating decoders that remain accurate at high erasure rates. We introduce PRISM, a transformer-based machine-learning decoder designed to exploit the known locations of erasures in protocols subject to both Pauli and erasure errors. We benchmark PRISM across a wide range of fault-tolerant protocols based on surface codes, color codes, and quantum low-density parity-check codes, with distances up to 24. Across these benchmarks, PRISM matches the logical error rates of integer-programming-based decoders while being significantly faster, including at error rates up to 50% of threshold. We further optimize the decoder and its real-time hardware implementation to reduce latency while maintaining accuracy. Our optimized implementation on an AMD VP1902 FPGA can decode logical blocks of distance 8 in both space and time, demonstrating a latency of 6.59 microseconds and an amortized decoding time of 1.56 microseconds per logical block is possible. These results show that machine-learning decoding can provide accurate, low-latency inference for erasure-dominated fault-tolerant quantum architectures.

quant-ph↗

Computing associators of endomorphism fusion categories

Many applications of fusion categories, particularly in physics, require the associators or $F$-symbols to be known explicitly. Finding these matrices typically involves solving vast systems of coupled polynomial equations in large numbers of variables. In this work, we present an algorithm that allows associator data for some category with unknown associator to be computed from a Morita equivalent category with known data. Given a module category over the latter, we utilize the representation theory of a module tube category, built from the known data, to compute this unknown associator data. When the input category is unitary, we discuss how to ensure the obtained data is also unitary. We provide several worked examples to illustrate this algorithm. In addition, we include several Mathematica files showing how the algorithm can be used to compute the data for the Haagerup category $\mathcal{H}_1$, whose data was previously unknown.

math.QA↗

Computing data for Levin-Wen with defects

We demonstrate how to do many computations for non-chiral topological phases with defects. These defects may be 1-dimensional domain walls or 0-dimensional point defects. Using $\operatorname{Vec}(S_3)$ as a guiding example, we demonstrate how domain wall fusion and associators can be computed using generalized tube algebra techniques. These domain walls can be both between distinct or identical phases. Additionally, we show how to compute all possible point defects, and the fusion and associator data of these. Worked examples, tabulated data and Mathematica code are provided.

quant-ph↗

Computing Defects Associated to Bounded Domain Wall Structures: The $\operatorname{Vec}(\mathbb{Z}/p\mathbb{Z})$ Case

A domain wall structure consists of a planar graph with faces labeled by fusion categories/topological phases. Edges are labeled by bimodules/domain walls. When the vertices are labeled by point defects we get a compound defect. We present an algorithm, called the domain wall structure algorithm, for computing the compound defect. We apply this algorithm to show that the \emph{bimodule associator}, related to the $O_3$ obstruction of [Etingof et al., Quantum Topol. 1, 209 (2010), arXiv:0909.3140], is trivial for all domain walls of $\operatorname{Vec}(\mathbb{Z}/p\mathbb{Z})$. In the language of this paper, the ground states of the Levin-Wen model are compound defects. We use this to define a generalization of the Levin-Wen model with domain walls and point defects. The domain wall structure algorithm can be used to compute the ground states of these generalized Levin-Wen type models.

math.QA↗

Fusing Binary Interface Defects in Topological Phases: The $\operatorname{Vec}(\mathbb{Z}/p\mathbb{Z})$ case

A binary interface defect is any interface between two (not necessarily invertible) domain walls. We compute all possible binary interface defects in Kitaev's $\mathbb{Z}/p\mathbb{Z}$ model and all possible fusions between them. Our methods can be applied to any Levin-Wen model. We also give physical interpretations for each of the defects in the $\mathbb{Z}/p\mathbb{Z}$ model. These physical interpretations provide a new graphical calculus which can be used to compute defect fusion.

math.QA↗

Deligne categories and representations of the infinite symmetric group

We establish a connection between two settings of representation stability for the symmetric groups $S_n$ over $\mathbb{C}$. One is the symmetric monoidal category ${\rm Rep}(S_{\infty})$ of algebraic representations of the infinite symmetric group $S_{\infty} = \bigcup_n S_n$, related to the theory of ${\bf FI}$-modules. The other is the family of rigid symmetric monoidal Deligne categories $\underline{\rm Rep}(S_t)$, $t \in \mathbb{C}$, together with their abelian versions $\underline{\rm Rep}^{ab}(S_t)$, constructed by Comes and Ostrik. We show that for any $t \in \mathbb{C}$ the natural functor ${\rm Rep}(S_{\infty}) \to \underline{\rm Rep}^{ab}(S_t)$ is an exact symmetric faithful monoidal functor, and compute its action on the simple representations of $S_{\infty}$. Considering the highest weight structure on $\underline{\rm Rep}^{ab}(S_t)$, we show that the image of any object of ${\rm Rep}(S_{\infty})$ has a filtration with standard objects in $\underline{\rm Rep}^{ab}(S_t)$. As a by-product of the proof, we give answers to the questions posed by P. Deligne concerning the cohomology of some complexes in the Deligne category $\underline{\rm Rep}(S_t)$, and their specializations at non-negative integers $n$.

math.RT↗

Domain walls in topological phases and the Brauer-Picard ring for $\operatorname{Vec}(\mathbb{Z}/p\mathbb{Z})$

We show how to calculate the relative tensor product of bimodule categories (not necessarily invertible) using ladder string diagrams. As an illustrative example, we compute the Brauer-Picard ring for the fusion category $\operatorname{Vec}(\mathbb{Z}/p\mathbb{Z})$. Moreover, we provide a physical interpretation of all indecomposable bimodule categories in terms of domain walls in the associated topological phase. We show how this interpretation can be used to compute the Brauer-Picard ring from a physical perspective.

math.QA↗

Eigenvalues of rotations and braids in spherical fusion categories

We give formulae for the multiplicities of eigenvalues of generalized rotation operators in terms of generalized Frobenius-Schur indicators in a semisimple spherical tensor category $\mathcal{C}$. In particular, this implies that the entire collection of rotation eigenvalues for a fusion category can be computed from the fusion rules and the traces of rotation at finitely many tensor powers. We also establish a rigidity property for FS indicators of fusion categories with a given fusion ring via Jones's theory of planar algebras. If $\mathcal{C}$ is also braided, these formulae yield the multiplicities of eigenvalues for a large class of braids in the associated braid group representations. When $\mathcal{C}$ is modular, this allows one to determine the eigenvalues and multiplicities of braids in terms of just the $S$ and $T$ matrices.

math.QA↗

Computing the Minimal Model for the Quantum Symmetric Algebra

In this note, we use some of the tensor categorial machinery developed by the quantum algebra community to study algebraic objects which appear in representation stability. In MR3430359, Sam and Snowden prove that the twisted commutative algebra Sym is Morita equivalent to the horizontal strip category. Their proof relies on a lemma proved by Olver in MR924166. We give a self contained proof that replaces Olver's lemma with information about the associator in the underlying category of polynomial GL(infty)-representations. In fact, we prove a quantum analogue of the theorem. The classical version follows by letting the parameter converge to 1.

math.RT↗

Noetherianity and rooted trees

Let T be the category whose objects are rooted trees and morphisms are order embeddings preserving the root. We prove that finitely generated representations of T are Noetherian using techniques developed by Sam and Snowden which generalize classical Grobner theory. The proof uses a relative version of Kruskals tree Theorem.

math.RT↗