BZPEERReading, trust and collaboration for research
Menu

Discover

GraphFollow connections around a paper, topic or saved view.

Workspaces

HomePick up where your research flow left off.InboxHandle requests, replies and notifications from one place.CollectionsCurate reading lists, evidence sets and shareable dossiers.ResearchSee your private provenance graph, trail and imports.CollaborationMove from discovery into focused discussion rooms.PublishDraft, preview and publish full-text research articles.

Network

ExpertsFind specialists worth following or asking for help.CommunitiesJoin research circles organized around shared work.PartnersSee external organizations active around the same topics.

Opportunities

ListingsBrowse roles, calls and collaborations with clearer fit signals.

Account

Log inReturn to your reading and collaboration workspace.Create accountStart saving work, following people and joining teams.
Log inCreate account

Researcher profile

Alex J. W. Thom

Alex J. W. Thom contributes to research discovery and scholarly infrastructure.

ResearcherAffiliation not importedOpen to collaborate
physics.chem-phcond-mat.str-elphysics.comp-phquant-phmath-phmath.MPphysics.atom-ph

Trust snapshot

Quick read

Trust 21 - EmergingVerification L1Unclaimed author
13works
0followers
7topics
4close collaborators

Actions

Decide how to stay connected

Follow researcher0

Identity and collaboration

How to connect with this researcher

Claiming links this public author record to a researcher profile and unlocks direct collaboration workflows.

Log in to claim

Direct collaboration

Open a focused conversation when the fit is right

Claim this author entity first to unlock direct invitations.

Research graph

See the researcher in context

Open full explorer

Inspect adjacent work, topics, institutions and collaborators without jumping out to a separate graph page.

Building this graph slice

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Published work

13 published item(s)

preprint2023arXiv

Global descriptors of a water molecule for machine learning of potential energy surfaces

Machine learning of multi-dimensional potential energy surfaces, from purely ab initio datasets, has seen substantial progress in the past years. Gaussian processes, a popular regression method, have been very successful at producing realistic potential energy surfaces from sparse datasets allowing to reduce the computational cost of highly accurate models. However, there are many choices one has to take in the design of the kernel and the feature space which can substantially change the performance and the characteristic of the model. In this study we explore different Gaussian processes set ups and, more specifically, how the permutational invariance of the target surface is controlled through the feature space selection.

0 citations0 reviewsNo code linkedOpen access
preprint2023arXiv

Optimised Morse transform of a Gaussian process feature space

Morse projections are well-known in chemistry and allow one, within a Morse potential approximation, to redefine the potential in a simple quadratic form. The latter, being a non-linear transform, is also very helpful for machine learning methods as they improve the performance of models by projecting the feature space onto more well-suited coordinates. Usually, the Morse projection parameters are taken from numerical benchmarks. We investigate the effect of changing these parameters latter on the model learning, as well as using the machine learning method itself to make the parameters decision. We find that learning is not necessarily improved by the latter and that general Morse projections are extremely susceptible to changes in the training data.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Modified noise kernels in Gaussian process modelling of energy surfaces

We explore the use of non homogenous noise kernels in Gaussian process modelling to improve the potential energy curve models describing stochastic electronic structure data. We use the same noise kernels on energy curves describing deterministic electronic structure data by creating non-homogenous noise model. We observe, as well as agreement between the noise of the model and the stochastic data, a strong regularisation of the curves when using artificial noises on deterministic data which improves Gaussian processes in the over fitting regime.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

On Symmetry and the Reality of Holomorphic Hartree--Fock Wavefunctions

The coalescence and disappearance of Hartree--Fock (HF) solutions as the molecular structure varies have been a common source of criticism for the breakdown of the HF approximation to the potential energy surfaces. However, recent developments in holomorphic HF theory show that this disappearing behavior is only a manifestation of the way conventional HF equations prevent solutions from being analytically continued, but it is unclear what factors govern the existence and the locations of these disappearances. In this work, we explore some of these factors from the perspective of spatial symmetry by introducing a classification for symmetry constraints on electronic-structure calculations. This forms a framework for us to systematically investigate several analytic holomorphic HF solutions of a model $\textrm{[H}_4\textrm{]}^{2+}$ system in STO-3G and demonstrate that, under appropriate conditions, spatial symmetry imposes strict requirements on the reality of certain solutions. The implications for self-consistent-field HF search algorithms are then discussed. Throughout this article, the term reality means the quality of a holomorphic HF solution having real molecular orbitals.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Reducing Unitary Coupled Cluster Circuit Depth by Classical Stochastic Amplitude Pre-Screening

Unitary Coupled Cluster (UCC) approaches are an appealing route to utilising quantum hardware to perform quantum chemistry calculations, as quantum computers can in principle perform UCC calculations in a polynomially scaling fashion, as compared to the exponential scaling required on classical computers. Current noisy intermediate scale quantum (NISQ) computers are limited by both hardware capacity in number of logical qubits and the noise introduced by the deep circuits required for UCC calculations using the Variational Quantum Eigensolver (VQE) approach. We present a combined classical--quantum approach where a stochastic classical UCC pre-processing step is used to determine the important excitations in the UCC ansatz. The reduced number of selected excitations are then used in a UCC-based VQE calculation. This approach gives a systematically improvable approximation, and we show that significant reductions in quantum resources can be achieved, with simulations on the CH$_2$, N$_2$ and N$_2$H$_2$ molecules giving sub-milliHartree errors.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

A Stochastic Approach to Unitary Coupled Cluster

Unitary coupled cluster (UCC), originally developed as a variational alternative to the popular traditional coupled cluster method, has seen a resurgence as a functional form for use on quantum computers. However, the number of excitors present in the ansatz often presents a barrier to implementation on quantum computers. Given the natural sparsity of wavefunctions obtained from Quantum Monte Carlo methods, we consider here a stochastic solution to the UCC problem. Using the Coupled Cluster Monte Carlo framework, we develop cluster selection schemes that capture the structure of the UCC wavefunction, as well as its Trotterized approximation, and use these to solve the corresponding projected equations. Due to the fast convergence of the equations with order in the cluster expansion, this approach scales polynomially with the size of the system. Unlike traditional UCC implementations, our approach naturally produces a non-variational estimator for the energy in the form of the projected energy. For UCCSD in small systems, we find this agrees well with the expectation value of the energy and, in the case of two electrons, with full configuration interaction results. For the larger N$_2$ system, the two estimators diverge, with the projected energy approaching the coupled cluster result, while the expectation value is close to results from traditional UCCSD.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Holomorphic Density Functional Theory

Self-consistent-field (SCF) approximations formulated using Hartree-Fock (HF) or Kohn-Sham Density Functional Theory (KS-DFT) both have the potential to yield multiple solutions. However, the formal relationship between multiple solutions identified using HF or KS-DFT remains generally unknown. We investigate the connection between multiple SCF solutions for HF or KS-DFT by introducing a parametrised functional that scales between the two representations. Using the hydrogen molecule and a model of electron transfer, we continuously map multiple solutions from the HF potential to a KS-DFT description. We discover that multiple solutions can coalesce and vanish as the functional changes, forming a direct analogy with the disappearance of real HF solutions along a change in molecular structure. To overcome this disappearance of solutions, we develop a complex-analytic extension of DFT - the "holomorphic DFT" approach - that allows every SCF stationary state to be analytically continued across all molecular structures and exchange-correlation functionals.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Reaching Full Correlation through Nonorthogonal Configuration Interaction: A Second-Order Perturbative Approach

Nonorthogonal multireference methods can predict statically correlated adiabatic energies while providing chemical insight through the combination of diabatic reference states. However, reaching quantitative accuracy using nonorthogonal multireference expansions remains a significant challenge. In this work, we present the first rigorous perturbative correction to nonorthogonal configuration interaction, allowing the remaining dynamic correlation to be reliably computed. Our second-order "NOCI-PT2" theory exploits a zeroth-order generalised Fock Hamiltonian and builds the first-order interacting space using single and double excitations from each reference determinant. This approach therefore defines the rigorous nonorthogonal extension to conventional multireference perturbation theory. We find that NOCI-PT2 can quantitatively predict multireference potential energy surfaces and provides state-specific ground and excited states for adiabatic avoided crossings. Furthermore, we introduce an explicit imaginary-shift formalism requiring shift values that are an order of magnitude smaller than those used in conventional multireference perturbation theory.

0 citations0 reviewsNo code linkedOpen access
preprint2019arXiv

$\mathcal{PT}$-Symmetry in Hartree-Fock Theory

$\mathcal{PT}$-symmetry --- invariance with respect to combined space reflection $\mathcal{P}$ and time reversal $\mathcal{T}$ --- provides a weaker condition than (Dirac) Hermiticity for ensuring a real energy spectrum of a general non-Hermitian Hamiltonian. $\mathcal{PT}$-symmetric Hamiltonians therefore form an intermediate class between Hermitian and non-Hermitian Hamiltonians. In this work, we derive the conditions for $\mathcal{PT}$-symmetry in the context of electronic structure theory, and specifically, within the Hartree-Fock (HF) approximation. We show that the HF orbitals are symmetric with respect to the $\mathcal{PT}$ operator \textit{if and only if} the effective Fock Hamiltonian is $\mathcal{PT}$-symmetric, and \textit{vice versa}. By extension, if an optimal self-consistent solution is invariant under $\mathcal{PT}$, then its eigenvalues and corresponding HF energy must be real. Moreover, we demonstrate how one can construct explicitly $\mathcal{PT}$-symmetric Slater determinants by forming $\mathcal{PT}$ doublets (i.e. pairing each occupied orbital with its $\mathcal{PT}$-transformed analogue), allowing $\mathcal{PT}$-symmetry to be conserved throughout the self-consistent process. Finally, considering the \ce{H2} molecule as an illustrative example, we observe $\mathcal{PT}$-symmetry in the HF energy landscape and find that the symmetry-broken unrestricted HF wave functions (i.e. diradical configurations) are $\mathcal{PT}$-symmetric, while the symmetry-broken restricted HF wave functions (i.e. ionic configurations) break $\mathcal{PT}$-symmetry.

0 citations0 reviewsNo code linkedOpen access
preprint2019arXiv

Accelerating Stochastic Quantum Chemistry

The convergence of full configuration interaction quantum Monte Carlo (FCIQMC) is accelerated using a quasi-Newton propagation (QN) which can also be applied to coupled cluster Monte Carlo (CCMC). The computational scaling of this optimised propagation is O(1), keeping the additional computational cost to a bare minimum. Its effects are investigated deterministically and stochastically on a model system, the uniform electron gas, with Hilbert space size up to $10^{40}$ and shown to accelerate convergence of the instantaneous projected energy by over an order of magnitude in the FCIQMC test case. Its capabilities are then demonstrated with FCIQMC on an archetypical quantum chemistry problem, the chromium dimer, in an all-electron basis set with Hilbert space size of about $10^{22}$ yielding highly accurate FCI energies.

0 citations0 reviewsNo code linkedOpen access
preprint2019arXiv

Complex Adiabatic Connection: a Hidden Non-Hermitian Path from Ground to Excited States

Processes related to electronically excited states are central in many areas of science, however accurately determining excited-state energies remains a major challenge in theoretical chemistry. Recently, higher energy stationary states of non-linear methods have themselves been proposed as approximations to excited states, although the general understanding of the nature of these solutions remains surprisingly limited. In this Letter, we present an entirely novel approach for exploring and obtaining excited stationary states by exploiting the properties of non-Hermitian Hamiltonians. Our key idea centres on performing analytic continuations of conventional quantum chemistry methods. Considering Hartree--Fock theory as an example, we analytically continue the electron-electron interaction to expose a hidden connectivity of multiple solutions across the complex plane, revealing a close resemblance between Coulson--Fischer points and non-Hermitian degeneracies. Finally, we demonstrate how a ground-state wave function can be morphed naturally into an excited-state wave function by constructing a well-defined complex adiabatic connection.

0 citations0 reviewsNo code linkedOpen access
preprint2019arXiv

Making the most of data: Quantum Monte Carlo Post-Analysis Revisited

In quantum Monte Carlo (QMC) methods, energy estimators are calculated as the statistical average of the Markov chain sampling of energy estimator along with an associated statistical error. This error estimation is not straightforward and there are several choices of the error estimation methods. We evaluate the performance of three methods, Straatsma, an autoregressive model, and a blocking analysis based on von Neumann's ratio test for randomness, for the energy time-series given by Diffusion Monte Carlo, Full Configuration Interaction Quantum Monte Carlo and Coupled Cluster Monte Carlo methods. From these analyses we describe a hybrid analysis method which provides reliable error estimates for series of all lengths. Equally important is the estimation of the appropriate start point of the equilibrated phase, and two heuristic schemes are tested, establishing that MSER (mean squared error rule) gives reasonable and constant estimations independent of the length of time-series.

0 citations0 reviewsNo code linkedOpen access
preprint2018arXiv

Exciting determinants in Quantum Monte Carlo: Loading the dice with fast, low memory weights

High-quality excitation generators are crucial to the effectiveness of Coupled cluster Monte Carlo (CCMC) and full configuration interaction Quantum Monte Carlo (FCIQMC) calculations. The heat bath sampling of Holmes et al. [A. A. Holmes, H. J. Changlani, and C. J. Umrigar, J. Chem. Theory Comput. 12, 1561 (2016)] dramatically increases the efficiency of the spawn step of such algorithms but requires memory storage scaling quartically with system size which can be prohibitive for large systems. Alavi et al. [S. D. Smart, G. H. Booth, and A. Alavi, unpublished] then approximated these weights with weights based on Cauchy--Schwarz-like inequalities calculated on-the-fly. While reducing the memory cost, this algorithm scales linearly in system size computationally. We combine both these ideas with the single reference nature of many systems, and introduce a spawn-sampling algorithm that has low memory requirements (quadratic in basis set size) compared to the heat bath algorithm and only scales either independently of system size (CCMC) or linearly in the number of electrons (FCIQMC). On small water chains with localized orbitals, we show that it is equally efficient as the other excitation generators. As the system gets larger, it converges faster than the on-the-fly weight algorithm, while having a much more favourable memory scaling than the heat bath algorithm.

0 citations0 reviewsNo code linkedOpen access