Home Research Notes Art

Jeffrey M. Epstein

I am primarily a quantum information theorist, and enjoy thinking about problems related to the information processing capabilities of quantum systems. This includes both work on the use of quantum systems for achieving information-processing tasks as well as work on the use of classical information (i.e. measurement outcomes) to infer models of quantum processes. During my PhD, I also worked on some problems in the relatively new field known as active matter, studying the emergent hydrodynamics of classical systems driven away from equilibrium at the single-particle level. Some selected publications from both of these areas are listed here; for a list of all publications, see my Google Scholar page.


Note on simple and consistent gateset characterization including calibration and decoherence errors.
JME. 2024 (arXiv)

This short paper describes a method for distinguishing X and Z basis decoherence channels acting as errors on X90 gates in a way that is robust to SPAM errors as well as over-rotation errors on the X90 gates. It also serves as a self-contained guide to performing efficient and informative characterization of simple Markovian errors in simple gatesets in a way that is fairly easy to interpret statistically.

Subspace Correction for Constraints.
KA Pawlak, JME, D Crow, S Gandhari, M Li, TC Bohdanowicz, J King. 2024 (arXiv)

We describe an interesting algorithmic primitive related to the implementation of constraints on a quantum computer, including a description of a protocol for the preparation of uniform superpositions over the set of all independent sets of a graph. Such a state preparation protocol could provide useful inputs to further computation.

Quantum noise limits for a class of nonlinear amplifiers.
JME, KB Whaley, J Combes. 2021 (arXiv)

This paper is a contribution to the quantum theory of measurement, demonstrating that projective measurements of arbitrary normal operators may be performed via nonlinear amplification.

Time reversal symmetry breaking in two-dimensional non-equilibrium viscous fluids.
JME, KK Mandadapu. 2020 (arXiv)

This is an analysis of the possible kinds of isotropic fluids in two dimensions, exploring the consequences of breaking equilibrium at the microscopic level. Using Onsager's regression hypothesis, we provide Green-Kubo formulae for the transport coefficients in terms of stress auto-correlation functions, showing that for systems without internal spin, the so-called odd viscosity requires breaking of time-reversal symmetry breaking at the microscopic level (as is present in some kinds of active matter). For systems with internal spin, there remains the possibility of non-vanishing odd viscosity even when time-reversal symmetry is preserved.

Time reversal symmetry breaking and odd viscosity in active fluids: Green-Kubo and NEMD results.
C Hargus, K Klymko, JME, KK Mandadapu. 2020 (arXiv)

This paper presents a numerical validation of the Green-Kubo relations for non-equilibrium viscosities derived in the above work via comparison to molecular dynamics simulations. More broadly, this provides some justification for the use of Onsager's regression hypothesis in the context of non-equilibrium statistical mechanics.

Quantum Speed Limits for Quantum Information Processing Tasks. JME, KB Whaley. 2017 (arXiv)

This work clarifies the connection between locality of dynamics in spin systems and fundamental speed limits on information processing tasks such as state transfer from one part of the system to another and generation of entanglement between distant regions. The basic tool of the analysis is the Lieb-Robinson bound.

Investigating the limits of randomized benchmarking protocols. JME, AW Cross, E Magesan, JM Gambetta. 2014 (arXiv)

This largely numerical study demonstrates the robustness of randomized benchmarking (RB) to various types of noise that violate the assumption of gate- and time-independent noise used in the (earliest) analysis of the method's statistical properties. As such it provides some validation for using RB to characterize the average gate fidelities of processors subject to more realistic noise.