aiqbench analysis
This paper advances quantum information theory by applying categorical methods to prove properties of generalized inverses for quantum channels, specifically highlighting the Drazin inverse's trace-preserving nature for error mitigation applications. The work bridges mathematical category theory with practical quantum computing challenges, offering rigorous theoretical foundations for handling non-invertible quantum operations.
Generalized Inverses of Quantum Channels: a categorical perspective
arXiv:2603.13946v2 Announce Type: replace Abstract: A quantum channel is defined as being completely positive (CP) and trace preserving (TP). While not every quantum channel is invertible or reversible, every quantum channel admits various kinds of generalized inverses such as the Moore-Penrose inverse and the Drazin inverse. A generalized inverse of a quantum channel may not itself be a quantum channel: it often fails to be CP. However, generalized inverses still play an important role in quantum error mitigation. Here, because it is often desirable for the generalized inverse of a quantum channel to be at least TP, the Drazin inverse, which is TP, is favoured over the Moore-Penrose inverse, which is not in general TP. In this paper, we take a categorical perspective on generalized inverses of quantum channels. This allows us to give a simple proof of the fact that the Drazin inverse of a quantum channel is always TP. It also allows us to show that for unital quantum channels, the Drazin inverse is also unital. We then generalize this result to dagger Drazin inverses, which allows us to show that for unital quantum channels, the Moore-Penrose inverse is both TP and unital as well. This opens the door to new applications of both the Drazin inverse and Moore-Penrose inverse in quantum information theory and, in particular, in quantum error mitigation.
Abstract:A quantum channel is defined as being completely positive (CP) and trace preserving (TP). While not every quantum channel is invertible or reversible, every quantum channel admits various kinds of generalized inverses such as the Moore-Penrose inverse and the Drazin inverse. A generalized inverse of a quantum channel may not itself be a quantum channel: it often fails to be CP. However, generalized inverses still play an important role in quantum error mitigation. Here, because it is often desirable for the generalized inverse of a quantum channel to be at least TP, the Drazin inverse, which is TP, is favoured over the Moore-Penrose inverse, which is not in general TP. In this paper, we take a categorical perspective on generalized inverses of quantum channels. This allows us to give a simple proof of the fact that the Drazin inverse of a quantum channel is always TP. It also allows us to show that for unital quantum channels, the Drazin inverse is also unital. We then generalize this result to dagger Drazin inverses, which allows us to show that for unital quantum channels, the Moore-Penrose inverse is both TP and unital as well. This opens the door to new applications of both the Drazin inverse and Moore-Penrose inverse in quantum information theory and, in particular, in quantum error mitigation.
In Proceedings QPL 2026, arXiv:2611.04069
Quantum Physics (quant-ph); Category Theory (math.CT)
arXiv:2603.13946 [quant-ph]
(or arXiv:2603.13946v2 [quant-ph] for this version)
https://doi.org/10.48550/arXiv.2603.13946 Focus to learn more arXiv-issued DOI via DataCite
EPTCS 456, 2026, pp. 128-143
https://doi.org/10.4204/EPTCS.456.4 Focus to learn more DOI(s) linking to related resources
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 333 words
Related coverage
Exact Functional ANOVA Decomposition for Categorical Inputs Models
arXiv · Jun 10, 2026
Foundations of Equivariant Deep Learning: Unifying Graph and Sheaf Neural Networks
arXiv · Aug 5, 2026
cGAP: Generalized Association Plots with HOMALS-Guided Heatmaps for Visualization of High-Dimensional Categorical Data
arXiv · Jul 20, 2026
Sparse Autoencoders Reveal Interpretable and Steerable Features in VLA Models
arXiv · Jun 9, 2026