Publications

Order by: publication date number of citations

• Generative learning of continuous data by tensor networks

  Alex Meiburg, Jing Chen, Jacob Miller, Raphaëlle Tihon, Guillaume Rabusseau, Alejandro Perdomo-Ortiz

  SciPost Phys. 18, 096 (2025) · published 17 March 2025 | Toggle abstract · pdf

  Beyond their origin in modeling many-body quantum systems, tensor networks have emerged as a promising class of models for solving machine learning problems, notably in unsupervised generative learning. While possessing many desirable features arising from their quantum-inspired nature, tensor network generative models have previously been largely restricted to binary or categorical data, limiting their utility in real-world modeling problems. We overcome this by introducing a new family of tensor network generative models for continuous data, which are capable of learning from distributions containing continuous random variables. We develop our method in the setting of matrix product states, first deriving a universal expressivity theorem proving the ability of this model family to approximate any reasonably smooth probability density function with arbitrary precision. We then benchmark the performance of this model on several synthetic and real-world datasets, finding that the model learns and generalizes well on distributions of continuous and discrete variables. We develop methods for modeling different data domains, and introduce a trainable compression layer which is found to increase model performance given limited memory or computational resources. Overall, our methods give important theoretical and empirical evidence of the efficacy of quantum-inspired methods for the rapidly growing field of generative learning.

• Space-time first-order correlations of an open Bose Hubbard model with incoherent pump and loss

  Martina Zündel, Leonardo Mazza, Léonie Canet, Anna Minguzzi

  SciPost Phys. 18, 095 (2025) · published 17 March 2025 | Toggle abstract · pdf

  We investigate the correlation properties in the steady state of driven-dissipative interacting bosonic systems in the quantum regime, as for example non-linear photonic cavities. Specifically, we consider the Bose-Hubbard model on a periodic chain and with spatially homogeneous one-body loss and pump within the Markovian approximation. The steady state is non-thermal and is formally equivalent to an infinite-temperature state with finite chemical potential set by the dissipative parameters. While there is no effect of interactions on the steady state, we observe a nontrivial behaviour of the space-time two-point correlation function, obtained by exact diagonalisation. In particular, we find that the decay width of the propagator is not only renormalised at increasing interactions, as it is the case of a single non-linear resonator, but also at increasing hopping strength. Furthermore, we numerically predict at large interactions a plateau value of the decay rate which goes beyond perturbative results in the interaction strength. We then compute the full spectral function, finding that it contains both a dispersive free-particle like dispersion at low energy and a doublon branch at energy corresponding to the on-site interactions. We compare with the corresponding calculation for the ground state of a closed quantum system and show that the driven-dissipative nature – determining both the steady state and the dynamical evolution – changes the low-lying part of the spectrum, where noticeably, the dispersion is quadratic instead of linear at small wavevectors. Finally, we compare to a high temperature grand-canonical equilibrium state and show the difference with respect to the open system stemming from the additional degree of freedom of the dissipation that allows one to vary the width of the dispersion lines.

• Pivoting through the chiral-clock family

  Nick G. Jones, Abhishodh Prakash, Paul Fendley

  SciPost Phys. 18, 094 (2025) · published 14 March 2025 | Toggle abstract · pdf

  The Onsager algebra, invented to solve the two-dimensional Ising model, can be used to construct conserved charges for a family of integrable $N$-state chiral clock models. We show how it naturally gives rise to a "pivot" procedure for this family of chiral Hamiltonians. These Hamiltonians have an anti-unitary CPT symmetry that when combined with the usual $\mathbb{Z}_N$ clock symmetry gives a non-Abelian dihedral symmetry group $D_{2N}$. We show that this symmetry gives rise to symmetry-protected topological (SPT) order in this family for all even $N$, and representation-SPT (RSPT) physics for all odd $N$. The simplest such example is a next-nearest-neighbour chain generalising the spin-1/2 cluster model, an SPT phase of matter. We derive a matrix-product state representation of its fixed-point ground state along with the ensuing entanglement spectrum and symmetry fractionalisation. We analyse a rich phase diagram combining this model with the Onsager-integrable chiral Potts chain, and find trivial, symmetry-breaking and (R)SPT orders, as well as extended gapless regions. For odd $N$, the phase transitions are "unnecessarily" critical from the SPT point of view.

• Higher-genus Fay-like identities from meromorphic generating functions

  Konstantin Baune, Johannes Broedel, Egor Im, Artyom Lisitsyn, Yannis Moeckli

  SciPost Phys. 18, 093 (2025) · published 14 March 2025 | Toggle abstract · pdf

  A possible way of constructing polylogarithms on Riemann surfaces of higher genera facilitates integration kernels, which can be derived from generating functions incorporating the geometry of the surface. Functional relations among polylogarithms rely on identities for those integration kernels. In this article, we derive identities for Enriquez' meromorphic generating function and investigate the implications for the associated integration kernels. The resulting identities are shown to be exhaustive and therefore reproduce all identities for Enriquez' kernels conjectured in arXiv:2407.11476 recently.

• Lefschetz thimble-inspired weight regularizations for complex Langevin simulations

  Kirill Boguslavski, Paul Hotzy, David I. Müller

  SciPost Phys. 18, 092 (2025) · published 14 March 2025 | Toggle abstract · pdf

  Complex Langevin (CL) is a computational method to circumvent the numerical sign problem with applications in finite-density quantum chromodynamics and the real-time dynamics of quantum field theories. It has long been known that, depending on the simulated system, CL does not always converge correctly. In this work, we provide numerical evidence that the success or failure of the complex Langevin method is deeply tied to the Lefschetz thimble structure of the simulated system. This is demonstrated by constructing weight function regularizations that deform the thimbles of systems with compact domains. Our results indicate that CL converges correctly when the regularized system exhibits a single relevant compact thimble. We introduce a bias correction to retrieve the values of the original theory for parameter sets where a direct complex Langevin approach fails. The effectiveness of this method is illustrated using several toy models, including the cosine model and the SU(2) and SU(3) Polyakov chains. Finally, we discuss the opportunities and limitations of this regularization approach for lattice field theories.

  1 citation

• Galilei particles revisited

  José Miguel Figueroa-O'Farrill, Simon Pekar, Alfredo Pérez, Stefan Prohazka

  SciPost Phys. Lect. Notes 93 (2025) · published 13 March 2025 | Toggle abstract · pdf

  We revisit the classifications of classical and quantum galilean particles: that is, we fully classify homogeneous symplectic manifolds and unitary irreducible projective representations of the Galilei group. Equivalently, these are coadjoint orbits and unitary irreducible representations of the Bargmann group, the universal central extension of the Galilei group. We provide an action principle in each case, discuss the nonrelativistic limit, as well as exhibit, whenever possible, the unitary irreducible representations in terms of fields on Galilei spacetime. Motivated by a forthcoming study of planons we pay close attention to the mobility of the less familiar massless Galilei particles.

• Codebase release 1.0 for Cytnx

  Kai-Hsin Wu, Chang-Teng Lin, Ke Hsu, Hao-Ti Hung, Manuel Schneider, Chia-Min Chung, Ying-Jer Kao, Pochung Chen

  SciPost Phys. Codebases 53-r1.0 (2025) · published 13 March 2025 | Toggle abstract · src

  We introduce a tensor network library designed for classical and quantum physics simulations called Cytnx (pronounced as sci-tens). This library provides an almost identical interface and syntax for both C++ and Python, allowing users to effortlessly switch between the two languages. Aiming at a quick learning process for new users of tensor network algorithms, the interfaces resemble the popular Python scientific libraries like NumPy, Scipy, and PyTorch. Not only multiple global Abelian symmetries can be easily defined and implemented, Cytnx also provides a new tool called Network that allows users to store large tensor networks and perform tensor network contractions in an optimal order automatically. With the integration of cuQuantum, tensor calculations can also be executed efficiently on GPUs. We present benchmark results for tensor operations on both devices, CPU and GPU. We also discuss features and higher-level interfaces to be added in the future.

  1 citation

• The Cytnx library for tensor networks

  Kai-Hsin Wu, Chang-Teng Lin, Ke Hsu, Hao-Ti Hung, Manuel Schneider, Chia-Min Chung, Ying-Jer Kao, Pochung Chen

  SciPost Phys. Codebases 53 (2025) · published 13 March 2025 | Toggle abstract · pdf

  We introduce a tensor network library designed for classical and quantum physics simulations called Cytnx (pronounced as sci-tens). This library provides an almost identical interface and syntax for both C++ and Python, allowing users to effortlessly switch between the two languages. Aiming at a quick learning process for new users of tensor network algorithms, the interfaces resemble the popular Python scientific libraries like NumPy, Scipy, and PyTorch. Not only multiple global Abelian symmetries can be easily defined and implemented, Cytnx also provides a new tool called Network that allows users to store large tensor networks and perform tensor network contractions in an optimal order automatically. With the integration of cuQuantum, tensor calculations can also be executed efficiently on GPUs. We present benchmark results for tensor operations on both devices, CPU and GPU. We also discuss features and higher-level interfaces to be added in the future.

  1 citation

• Scattering from an external field in quantum chromodynamics at high energies: From foundations to interdisciplinary connections

  Athanasia-Konstantina Angelopoulou, Anh Dung Le, Stéphane Munier

  SciPost Phys. Lect. Notes 92 (2025) · published 12 March 2025 | Toggle abstract · pdf

  We review the factorization of the $S$-matrix elements in the context of particle scattering off an external field, which can serve as a model for the field of a large nucleus. The factorization takes the form of a convolution of light cone wave functions describing the physical incoming and outgoing states in terms of bare partons, and products of Wilson lines. The latter represent the interaction between the bare partons and the external field. Specializing to elastic scattering amplitudes of onia at very high energies, we introduce the color dipole model, which formulates the calculation of the modulus-squared of the wave functions in quantum chromodynamics with the help of a branching random walk, and the scattering amplitudes as observables on this classical stochastic process. Methods developed for general branching processes produce analytical formulas for the asymptotics of such observables, and thus enable one to derive exact large-rapidity expressions for onium-nucleus cross sections, from which electron-nucleus cross sections may be inferred.

• Nonlinearity of transparent SNS weak links decreases sharply with length

  Valla Fatemi, Pavel D. Kurilovich, Anton R. Akhmerov, Bernard van Heck

  SciPost Phys. 18, 091 (2025) · published 12 March 2025 | Toggle abstract · pdf

  Superconductor-normal material-superconductor (SNS) junctions are being integrated into microwave circuits for fundamental and applied research goals. The short junction limit is a common simplifying assumption for experiments with SNS junctions, but this limit constrains how small the nonlinearity of the microwave circuit can be. Here, we show that a finite length of the weak link strongly suppresses the nonlinearity compared to its zero-length limit -- the suppression can be up to a factor of ten even when the length remains shorter than the induced coherence length. We tie this behavior to the nonanalytic dependence of nonlinearity on length, which the critical current does not exhibit. Further, we identify additional experimentally observable consequences of nonzero length, and we conjecture that anharmonicity is bounded between zero and a maximally negative value for any non-interacting Josephson junction in the presence of time-reversal symmetry. We promote SNS junction length as a useful parameter for designing weakly nonlinear microwave circuits.