Georgios Korpas
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Georgios Korpas has a total of 67 co-authors. Here are the top co-authors by number of shared publications:
- Jakub Marecek
Czech Technical University in Prague
4 shared publications - Jan Manschot
School of Mathematics, Trinity College, Dublin, Ireland
4 shared publications - Johannes Aspman
Czech Technical University in Prague
3 shared publications - Vyacheslav Kungurtsev
Department of Computer Science Czech Technical University Karlovo Namesti 13 121 35 Prague 2 Czech Republic
3 shared publications - Elias Furrer
University of Birmingham
2 shared publications - Gregory W. Moore
NHETC and Department of Physics and Astronomy, Rutgers University, Piscataway, USA
2 shared publications
Latest Publications
- Open Access
- 11 April 2026
- Published by Elsevier in SSRN Electronic Journal
Federated Sinkhorn: Communication, Performance, and Privacy for Distributed Entropic Optimal Transport
We study distributed Sinkhorn iterations for entropy-regularized optimal transport when the Gibbs kernel operator is row-partitioned across c workers and cannot be centralized. We present Federated Sinkhorn, two exact synchronous protocols that exchange only scaling-vector slices: (i) an All-to-All scheme implemented by Allgather, and (ii) a Star (parameter-server) scheme implemented by client-to-server sends and server-to-client broadcasts. For both, we derive closed-form per-iteration compute, communication, and memory costs under an α-β latency--bandwidth model, and show that the distributed iterates match centralized Sinkhorn under standard positivity assumptions. Multi-node CPU/GPU experiments validate the model and show that repeated global scaling exchange quickly becomes the dominant bottleneck as c increases. We also report an optional bounded-delay asynchronous schedule and an optional privacy measurement layer for communicated log-scalings.
- 11 April 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessFederated Sinkhorn: Communication, Performance, and Privacy for Distributed Entropic Optimal Transport
We study distributed Sinkhorn iterations for entropy-regularized optimal transport when the Gibbs kernel operator is row-partitioned across c workers and cannot be centralized. We present Federated Sinkhorn, two exact synchronous protocols that exchange only scaling-vector slices: (i) an All-to-All scheme implemented by Allgather, and (ii) a Star (parameter-server) scheme implemented by client-to-server sends and server-to-client broadcasts. For both, we derive closed-form per-iteration compute, communication, and memory costs under an α-β latency--bandwidth model, and show that the distributed iterates match centralized Sinkhorn under standard positivity assumptions. Multi-node CPU/GPU experiments validate the model and show that repeated global scaling exchange quickly becomes the dominant bottleneck as c increases. We also report an optional bounded-delay asynchronous schedule and an optional privacy measurement layer for communicated log-scalings.
- Open Access
- 21 October 2025
- Published by IOS Press
- p. 2017-2024
Binarizing Physics-Inspired GNNs for Combinatorial Optimization
Physics-inspired graph neural networks (PI-GNNs) have been utilized as an efficient unsupervised framework for relaxing combinatorial optimization problems encoded through a specific graph structure and loss, reflecting dependencies between the problem’s variables. While the framework has yielded promising results in various combinatorial problems, we show that the performance of PI-GNNs systematically plummets with an increasing density of the combinatorial problem graphs. Our analysis reveals an interesting phase transition in the PI-GNNs’ training dynamics, associated with degenerate solutions for the denser problems, highlighting a discrepancy between the relaxed, real-valued model outputs and the binary-valued problem solutions. To address the discrepancy, we propose principled alternatives to the naive strategy used in PI-GNNs by building on insights from fuzzy logic and binarized neural networks. Our experiments demonstrate that the portfolio of proposed methods significantly improves the performance of PI-GNNs in increasingly dense settings.
- 21 October 2025
- Published by IOS Press
- p. 2017-2024
Open AccessBinarizing Physics-Inspired GNNs for Combinatorial Optimization
Physics-inspired graph neural networks (PI-GNNs) have been utilized as an efficient unsupervised framework for relaxing combinatorial optimization problems encoded through a specific graph structure and loss, reflecting dependencies between the problem’s variables. While the framework has yielded promising results in various combinatorial problems, we show that the performance of PI-GNNs systematically plummets with an increasing density of the combinatorial problem graphs. Our analysis reveals an interesting phase transition in the PI-GNNs’ training dynamics, associated with degenerate solutions for the denser problems, highlighting a discrepancy between the relaxed, real-valued model outputs and the binary-valued problem solutions. To address the discrepancy, we propose principled alternatives to the naive strategy used in PI-GNNs by building on insights from fuzzy logic and binarized neural networks. Our experiments demonstrate that the portfolio of proposed methods significantly improves the performance of PI-GNNs in increasingly dense settings.
- Top Cited Papers
- 28 October 2024
- Published by Springer Nature in Nature Reviews Physics
- Vol. 6 (12) , 718-735
Challenges and opportunities in quantum optimization
- 28 October 2024
- Published by Springer Nature in Nature Reviews Physics
- Vol. 6 (12) , 718-735
Top Cited PapersChallenges and opportunities in quantum optimization
- Open Access
- 10 October 2024
- Published by Verein zur Forderung des Open Access Publizierens in den Quantenwissenschaften in Quantum
- Vol. 8, 1495
Iteration Complexity of Variational Quantum Algorithms
There has been much recent interest in near-term applications of quantum computers, i.e., using quantum circuits that have short decoherence times due to hardware limitations. Variational quantum algorithms (VQA), wherein an optimization algorithm implemented on a classical computer evaluates a parametrized quantum circuit as an objective function, are a leading framework in this space. An enormous breadth of algorithms in this framework have been proposed for solving a range of problems in machine learning, forecasting, applied physics, and combinatorial optimization, among others.In this paper, we analyze the iteration complexity of VQA, that is, the number of steps that VQA requires until its iterates satisfy a surrogate measure of optimality. We argue that although VQA procedures incorporate algorithms that can, in the idealized case, be modeled as classic procedures in the optimization literature, the particular nature of noise in near-term devices invalidates the claim of applicability of off-the-shelf analyses of these algorithms. Specifically, noise makes the evaluations of the objective function via quantum circuits b i a s e d . Commonly used optimization procedures, such as SPSA and the parameter shift rule, can thus be seen as derivative-free optimization algorithms with biased function evaluations, for which there are currently no iteration complexity guarantees in the literature. We derive the missing guarantees and find that the rate of convergence is unaffected. However, the level of bias contributes unfavorably to both the constant therein, and the asymptotic distance to stationarity, i.e., the more bias, the farther one is guaranteed, at best, to reach a stationary point of the VQA objective.
- 10 October 2024
- Published by Verein zur Forderung des Open Access Publizierens in den Quantenwissenschaften in Quantum
- Vol. 8, 1495
Open AccessIteration Complexity of Variational Quantum Algorithms
There has been much recent interest in near-term applications of quantum computers, i.e., using quantum circuits that have short decoherence times due to hardware limitations. Variational quantum algorithms (VQA), wherein an optimization algorithm implemented on a classical computer evaluates a parametrized quantum circuit as an objective function, are a leading framework in this space. An enormous breadth of algorithms in this framework have been proposed for solving a range of problems in machine learning, forecasting, applied physics, and combinatorial optimization, among others.In this paper, we analyze the iteration complexity of VQA, that is, the number of steps that VQA requires until its iterates satisfy a surrogate measure of optimality. We argue that although VQA procedures incorporate algorithms that can, in the idealized case, be modeled as classic procedures in the optimization literature, the particular nature of noise in near-term devices invalidates the claim of applicability of off-the-shelf analyses of these algorithms. Specifically, noise makes the evaluations of the objective function via quantum circuits b i a s e d . Commonly used optimization procedures, such as SPSA and the parameter shift rule, can thus be seen as derivative-free optimization algorithms with biased function evaluations, for which there are currently no iteration complexity guarantees in the literature. We derive the missing guarantees and find that the rate of convergence is unaffected. However, the level of bias contributes unfavorably to both the constant therein, and the asymptotic distance to stationarity, i.e., the more bias, the farther one is guaranteed, at best, to reach a stationary point of the VQA objective.
- Open Access
- 24 March 2024
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence
- Vol. 38 (10) , 10935-10943
Taming Binarized Neural Networks and Mixed-Integer Programs
There has been a great deal of recent interest in binarized neural networks, especially because of their explainability. At the same time, automatic differentiation algorithms such as backpropagation fail for binarized neural networks, which limits their applicability. We show that binarized neural networks admit a tame representation by reformulating the problem of training binarized neural networks as a subadditive dual of a mixed-integer program, which we show to have nice properties. This makes it possible to use the framework of Bolte et al. for implicit differentiation, which offers the possibility for practical implementation of backpropagation in the context of binarized neural networks. This approach could also be used for a broader class of mixed-integer programs, beyond the training of binarized neural networks, as encountered in symbolic approaches to AI and beyond.
- 24 March 2024
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence
- Vol. 38 (10) , 10935-10943
Open AccessTaming Binarized Neural Networks and Mixed-Integer Programs
There has been a great deal of recent interest in binarized neural networks, especially because of their explainability. At the same time, automatic differentiation algorithms such as backpropagation fail for binarized neural networks, which limits their applicability. We show that binarized neural networks admit a tame representation by reformulating the problem of training binarized neural networks as a subadditive dual of a mixed-integer program, which we show to have nice properties. This makes it possible to use the framework of Bolte et al. for implicit differentiation, which offers the possibility for practical implementation of backpropagation in the context of binarized neural networks. This approach could also be used for a broader class of mixed-integer programs, beyond the training of binarized neural networks, as encountered in symbolic approaches to AI and beyond.
- Open Access
- 13 September 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Quantum Engineering
- Vol. 4, 1-18
Approaching Collateral Optimization for NISQ and Quantum-Inspired Computing (May 2023)
Collateral optimization refers to the systematic allocation of financial assets to satisfy obligations or secure transactions, while simultaneously minimizing costs and optimizing the usage of available resources. This involves assessing number of characteristics, such as cost of funding and quality of the underlying assets to ascertain the optimal collateral quantity to be posted to cover exposure arising from a given transaction or a set of transactions. One of the common objectives is to minimise the cost of collateral required to mitigate the risk associated with a particular transaction or a portfolio of transactions while ensuring sufficient protection for the involved parties. Often, this results in a large-scale combinatorial optimization problem. In this study, we initially present a Mixed Integer Linear Programming (MILP) formulation for the collateral optimization problem, followed by a Quadratic Unconstrained Binary optimization (QUBO) formulation in order to pave the way towards approaching the problem in a hybrid-quantum and NISQ-ready way. We conduct local computational small-scale tests using various Software Development Kits (SDKs) and discuss the behavior of our formulations as well as the potential for performance enhancements. We find that while the QUBO based approaches fail to find the global optima in the small scale experiments, they are reasonably close suggesting their potential for large instances. We further survey the recent literature that proposes alternative ways to attack combinatorial optimization problems suitable for collateral optimization.
- 13 September 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Quantum Engineering
- Vol. 4, 1-18
Open AccessApproaching Collateral Optimization for NISQ and Quantum-Inspired Computing (May 2023)
Collateral optimization refers to the systematic allocation of financial assets to satisfy obligations or secure transactions, while simultaneously minimizing costs and optimizing the usage of available resources. This involves assessing number of characteristics, such as cost of funding and quality of the underlying assets to ascertain the optimal collateral quantity to be posted to cover exposure arising from a given transaction or a set of transactions. One of the common objectives is to minimise the cost of collateral required to mitigate the risk associated with a particular transaction or a portfolio of transactions while ensuring sufficient protection for the involved parties. Often, this results in a large-scale combinatorial optimization problem. In this study, we initially present a Mixed Integer Linear Programming (MILP) formulation for the collateral optimization problem, followed by a Quadratic Unconstrained Binary optimization (QUBO) formulation in order to pave the way towards approaching the problem in a hybrid-quantum and NISQ-ready way. We conduct local computational small-scale tests using various Software Development Kits (SDKs) and discuss the behavior of our formulations as well as the potential for performance enhancements. We find that while the QUBO based approaches fail to find the global optima in the small scale experiments, they are reasonably close suggesting their potential for large instances. We further survey the recent literature that proposes alternative ways to attack combinatorial optimization problems suitable for collateral optimization.
- Open Access
Renormalization and BRST symmetry in Donaldson-Witten theory
The presence of a BRST symmetry in topologically twisted gauge theories makes a precise analysis of these theories feasible. While the global BRST symmetry suggests that correlation functions of BRST exact observables vanish, this decoupling might be obstructed due to a contribution from the boundary of field space. Motivated by divergent BRST exact observables on the Coulomb branch of Donaldson-Witten theory, we put forward a new prescription for the renormalization of correlation functions on the Coulomb branch. This renormalization is based on the relation between Coulomb branch integrals and integrals over a modular fundamental domain, and establishes that BRST exact observables indeed decouple in Donaldson-Witten theory.
Open AccessRenormalization and BRST symmetry in Donaldson-Witten theory
The presence of a BRST symmetry in topologically twisted gauge theories makes a precise analysis of these theories feasible. While the global BRST symmetry suggests that correlation functions of BRST exact observables vanish, this decoupling might be obstructed due to a contribution from the boundary of field space. Motivated by divergent BRST exact observables on the Coulomb branch of Donaldson-Witten theory, we put forward a new prescription for the renormalization of correlation functions on the Coulomb branch. This renormalization is based on the relation between Coulomb branch integrals and integrals over a modular fundamental domain, and establishes that BRST exact observables indeed decouple in Donaldson-Witten theory.
- Open Access
Globally Optimal Quantum Control
Optimization methods for constrained quantum control problems power quantum technologies. Such control problems are notoriously difficult because they are non-convex and plagued with local extrema. Current optimization methods must be repeated many times to find good solutions, each time requiring many simulations of the system. Here we present Quantum Control via Polynomial Optimization (QCPOp), a method that eliminates this problem by directly finding globally optimal solutions. This remarkable ability is due to global optimization methods recently developed for polynomial functions. We demonstrate the tremendous improvement over current state-of-the-art methods using a number of non-trivial examples. Global optimization also allows QCPOp to find the simplest control solutions. Since QCPOp is able to reveal the optimum performance of quantum control with high confidence, we expect that it will not only enhance the utility of such control but provide a key tool for determining the limits of quantum technologies.
Open AccessGlobally Optimal Quantum Control
Optimization methods for constrained quantum control problems power quantum technologies. Such control problems are notoriously difficult because they are non-convex and plagued with local extrema. Current optimization methods must be repeated many times to find good solutions, each time requiring many simulations of the system. Here we present Quantum Control via Polynomial Optimization (QCPOp), a method that eliminates this problem by directly finding globally optimal solutions. This remarkable ability is due to global optimization methods recently developed for polynomial functions. We demonstrate the tremendous improvement over current state-of-the-art methods using a number of non-trivial examples. Global optimization also allows QCPOp to find the simplest control solutions. Since QCPOp is able to reveal the optimum performance of quantum control with high confidence, we expect that it will not only enhance the utility of such control but provide a key tool for determining the limits of quantum technologies.
- Open Access
Quantum State Tomography as a Bilevel Problem, Utilizing I-Q Plane Data
It is natural to ask how to utilize actual measurements, such as the so-called IQ-plane data obtained in the dispersive readout of transmon qubits, in the estimation of the state of a quantum system. We formulate the joint problem of discrimination and quantum state tomography as a bilevel optimization problem and show how to solve it. The use of the joint problem can improve the sample complexity (or the reconstruction error for a fixed number of measurements) compared with traditional techniques that decompose the problem into the discrimination and state tomography based on the estimated expectation values of certain projective measurement operators.
Open AccessQuantum State Tomography as a Bilevel Problem, Utilizing I-Q Plane Data
It is natural to ask how to utilize actual measurements, such as the so-called IQ-plane data obtained in the dispersive readout of transmon qubits, in the estimation of the state of a quantum system. We formulate the joint problem of discrimination and quantum state tomography as a bilevel optimization problem and show how to solve it. The use of the joint problem can improve the sample complexity (or the reconstruction error for a fixed number of measurements) compared with traditional techniques that decompose the problem into the discrimination and state tomography based on the estimated expectation values of certain projective measurement operators.
- Open Access
- 26 March 2022
- Published by Springer Nature in Letters in Mathematical Physics
- Vol. 112 (2) , 1-53
The u-plane integral, mock modularity and enumerative geometry
We revisit the low-energy effective U(1) action of topologically twisted $${\mathcal {N}}=2$$ SYM theory with gauge group of rank one on a generic oriented smooth four-manifold X with nontrivial fundamental group. After including a specific new set of $$\mathcal Q$$ -exact operators to the known action, we express the integrand of the path integral of the low-energy U(1) theory as an anti-holomorphic derivative. This allows us to use the theory of mock modular forms and indefinite theta functions for the explicit evaluation of correlation functions of the theory, thus facilitating the computations compared to previously used methods. As an explicit check of our results, we compute the path integral for the product ruled surfaces $$X=\Sigma _g \times \mathbb {CP}^1$$ for the reduction on either factor and compare the results with existing literature. In the case of reduction on the Riemann surface $$\Sigma _g$$ , via an equivalent topological A-model on $$\mathbb {CP}^1$$ , we will be able to express the generating function of genus zero Gromov–Witten invariants of the moduli space of flat rank one connections over $$\Sigma _g$$ in terms of an indefinite theta function, whence we would be able to make concrete numerical predictions of these enumerative invariants in terms of modular data, thereby allowing us to derive results in enumerative geometry from number theory.- 26 March 2022
- Published by Springer Nature in Letters in Mathematical Physics
- Vol. 112 (2) , 1-53
Open AccessThe u-plane integral, mock modularity and enumerative geometry
We revisit the low-energy effective U(1) action of topologically twisted $${\mathcal {N}}=2$$ SYM theory with gauge group of rank one on a generic oriented smooth four-manifold X with nontrivial fundamental group. After including a specific new set of $$\mathcal Q$$ -exact operators to the known action, we express the integrand of the path integral of the low-energy U(1) theory as an anti-holomorphic derivative. This allows us to use the theory of mock modular forms and indefinite theta functions for the explicit evaluation of correlation functions of the theory, thus facilitating the computations compared to previously used methods. As an explicit check of our results, we compute the path integral for the product ruled surfaces $$X=\Sigma _g \times \mathbb {CP}^1$$ for the reduction on either factor and compare the results with existing literature. In the case of reduction on the Riemann surface $$\Sigma _g$$ , via an equivalent topological A-model on $$\mathbb {CP}^1$$ , we will be able to express the generating function of genus zero Gromov–Witten invariants of the moduli space of flat rank one connections over $$\Sigma _g$$ in terms of an indefinite theta function, whence we would be able to make concrete numerical predictions of these enumerative invariants in terms of modular data, thereby allowing us to derive results in enumerative geometry from number theory.