Quantum Cross Entropy in Integrated Fluctuation Theorems
ORAL
Abstract
We discuss the role of quantum cross entropy from the perspective of integrated fluctuation
theorems beyond the standard two-time measurement scheme. We employ the one-time
measurement scheme to construct an entropy gain distribution which is conditioned on the
outcome of an initial projective measurement. Then, we demonstrate that lower bounds on the
total entropy gain are tighter than those derived from the two-time measurement scheme and, in
turn, yield upper bounds on the classical information lost over a noisy quantum channel. Further
discussion applies fluctuation theorems to quantum machine learning and the thermodynamics of
a system undergoing heat emission.
theorems beyond the standard two-time measurement scheme. We employ the one-time
measurement scheme to construct an entropy gain distribution which is conditioned on the
outcome of an initial projective measurement. Then, we demonstrate that lower bounds on the
total entropy gain are tighter than those derived from the two-time measurement scheme and, in
turn, yield upper bounds on the classical information lost over a noisy quantum channel. Further
discussion applies fluctuation theorems to quantum machine learning and the thermodynamics of
a system undergoing heat emission.
*This work is supported by NSF: Proposal 2036347 (STTR Phase 1). A.S. was supported by the internal R&D from Aliro Technologies, Inc. Now, he is supported by the startup package from University of Massachusetts Boston. N.Y. is supported by MEXT Quantum Leap Flagship Program Grants No. JPMXS0118067285 and 9 No. JPMXS0120319794. T.J.H is supported by the graduate study program at University of Massachusetts Boston. Work by P.N. is supported by the NSF RAISE-QAC-QSA, Grant No. DMR-2037783 and the Department of Energy, Office of Basic Energy Sciences Grant DE-SC0019215.
–
Publication: arXiv:2209.01761 [quant-ph]
Presenters
-
Tharon Holdsworth
- University of Massachusetts Boston