Completed Theses
Doctoral Theses
- Michael Zurel, Classical descriptions of quantum computations : foundations of quantum computation via hidden variable models, quasiprobability representations, and classical simulation algorithms, 2024 (supervised by Robert Raußendorf)
- Oleg Kabernik, Reductions in finite-dimensional quantum mechanics : from symmetries to operator algebras and beyond, 2021 (supervised by Robert Raußendorf)
Master's Theses
- Ryohei Weil, Quantifying resource states and efficient regimes of measurement-based quantum computation on a superconducting processor, 2024 (supervised by Robert Raußendorf)
- Luis Mantilla Calderon, Measurement-based quantum machine learning, 2023 (supervised by Dmytro Bondarenko, Polina Feldmann, Robert Raußendorf)
- Arnab Adhikary, Symmetry protected measurement-based quantum computation in finite spin chains, 2021 (supervised by Robert Raußendorf)
- Paul Herringer, Classification of quantum wire in tensor network states with a local Pauli symmetry, 2021 (supervised by Robert Raußendorf)
- Michael Zurel, Hidden variable models and classical simulation algorithms for quantum computation with magic states on qubits, 2020 (supervised by Robert Raußendorf)
- David Stephen, Computational power of one-dimensional symmetry-protected topological phases, 2017 (supervised by Robert Raußendorf)
Bachelor's Theses
- Elektra Dakogiannis, Coherent Noise Models in Concatenated Stabilizer Codes, 2023 (supervised by Gabrielle Tournaire, Robert Raußendorf)
- Julian Ding, Simulation of the Kitaev planar code error threshold under a photonic local error model, 2022 (supervised by Robert Raußendorf)
- Ryohei Weil, A Simulation of a Simulation : Algorithms for Symmetry-Protected Measurement-Based Quantum Computing Experiments, 2022 (supervised by Robert Raußendorf)
- Aslan Zhang, Compilation Optimization for Measurement-based Quantum Computation on Graph States, 2022 (supervised by Polina Feldmann, Robert Raußendorf)
Current Topics
Doctoral Theses
- Solving separability problems with machine learning (contact: Martin Plávala)
Detecting quantum entanglement is a hard task: while some computable criteria exists, they may not be sufficient or may be impossible to compute beyond the smallest local dimensions of the underlying Hilbert spaces. A potential solution is offered by machine learning where various methods were used to classify separable and entangled states. These methods have their own pitfalls: they usually only predict whether the state is separable or entangled with some probability, but there is no way to verify this prediction or to estimate the noise-robustness of the entanglement.
This project aims to rectify all of these problems: the aim is to construct entanglement witnesses from a large class corresponding to a fixed level of the hierarchy of symmetric extensions; these entanglement witnesses are known to perform reasonably well in applications and they have favorable computational properties in that they can be numerically verified to be entanglement witnesses. Deep neural networks will be used to construct these entanglement witnesses; since the neural network will be guessing entanglement witness from the specific class, the result will be verifiable in that if entanglement is detected, explicit witness will be provided as well. The goal of the project is to make these methods work for high-dimensional cases where other methods fail due to a lack of computational resources.
As a follow-up project, the developed methods will be applied to high-dimensional steering and Bell nonlocality, since both of these problems are known to be analogical to separability and the developed methods will directly generalize to these cases. Further projects will include improving the performance of the machine learning methods and investigating nonlocality in the triangle network, the later will present a novel set of challenges since the underlying problem is not linear (unlike all of the previous problems).
The ideal candidate should have background in linear algebra and interest in quantum information, previous experience in machine learning and deep neural networks is welcome. The project will involve mainly developing and testing various machine learning models constructing the entanglement witnesses, the necessary mathematical calculations will likely involve only undergraduate level linear algebra and quantum information and relevant numerical methods. Thus all candidates should have interest in programming and numerical mathematics.
Master's Theses
-
k-compatibility of channels in general probabilistic theories (contact: Martin Plávala)
General probabilistic theories (GPTs) are a class of theories that include both classical and quantum information as well as other mathematically consistent theories. GPTs are a point of interest in quantum foundations when attempting to derive quantum theory, but they are also used in quantum information to reformulate and solve problems. A channel is a transformation of the states of the system, such as for example time evolution for a given amount of time. A set of channels is k-compatible if given k exact copies of the input state, there exist another channel that outputs a state that is locally identical with outputs of all of the considered channels. The goal of the project is to investigate when a set of channels is k-compatible within a given GPT, special attention will be put on connecting k-compatibility of several instances of the identify channel and geometric properties of the considered state space. The project will mostly involve only analytic methods and will be close to mathematical research.
-
Products of POVMs (contact: Martin Plávala)
POVMs are a mathematical representation of the measurement devices in quantum information. Most notably a set of POVMs can be compatible or jointly measurable, meaning that all of the measurements represented by the respective POVMs can be reduced to only a single measurement. The goal of the project is to investigate when for compatible POVMs the single measurement can be constructed using some notion of operator product, what are the properties of such operator products and whether they have any operational meaning. The project will mostly involve analytic methods with some numerics to construct appropriate examples.
Bachelor's Theses
Current Theses
Doctoral Theses
- Arnab Adhikary, Computational phases of quantum matter — Korrelationen verstehen (supervised by Robert Raußendorf)
- Ruben Campos Delgado, Quantenfehlertoleranz — Effiziente Quantencodes (supervised by Robert Raußendorf)
- Poya Haghnegahdar, Messungsbasiertes Quantenrechnen auf Affleck-Kennedy-Lieb-Tasaki Zuständen (supervised by Robert Raußendorf)
- Lukas Hantzko, Computational phases of quantum matter — Algebraische Struktur (supervised by Robert Raußendorf)
- Paul Herringer, Computational phases of quantum matter — Verbindungen mit der Physik kondensierter Materie (supervised by Robert Raußendorf)
- Thierry Kaldenbach, Graphenzustände und Quantenalgorithmen (supervised by Robert Raußendorf)
- Marvin Schwiering, Quantenrechner—Architektur (supervised by Robert Raußendorf)
- Gabrielle Tournaire, Fehlertolerantes Quantenrechen mit topologischen Quantencodes (supervised by Robert Raußendorf)
Master's Theses
- Niko Trittschanke, Handling Quantum Errors under Realistic Noise Models for Trapped Ion Quantum Devices (supervised by Robert Raußendorf)
Bachelor's Theses
