Unsupervised learning of quantum many-body scars using intrinsic dimension

Harvey Cao,Dimitris G. Angelakis,Daniel Leykam

2024-02-01

Abstract:Quantum many-body scarred systems contain both thermal and non-thermal scar eigenstates in their spectra. When these systems are quenched from special initial states which share high overlap with scar eigenstates, the system undergoes dynamics with atypically slow relaxation and periodic revival. This scarring phenomenon poses a potential avenue for circumventing decoherence in various quantum engineering applications. Given access to an unknown scar system, current approaches for identification of special states leading to non-thermal dynamics rely on costly measures such as entanglement entropy. In this work, we show how two dimensionality reduction techniques, multidimensional scaling and intrinsic dimension estimation, can be used to learn structural properties of dynamics in the PXP model and distinguish between thermal and scar initial states. The latter method is shown to be robust against limited sample sizes and experimental measurement errors.

Quantum Physics,Statistical Mechanics,Chaotic Dynamics

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to identify the special initial states that lead to non - thermal dynamic behaviors in quantum many - body systems. Specifically, the paper focuses on the problems in Quantum Many - Body Scars (QMBS) systems. These systems contain thermal and non - thermal scar eigenstates. When evolving from specific initial states with high overlap with the scar eigenstates, the system will exhibit an abnormally slow relaxation process and periodic recurrence phenomena. This scar phenomenon has potential application value in quantum engineering applications because it can overcome the decoherence problem.

The current methods for identifying these special initial states rely on costly metrics, such as entanglement entropy. These methods are difficult to implement in experiments, especially when the system scale increases. Therefore, the paper proposes a new method, using two dimension - reduction techniques - Multidimensional Scaling (MDS) and Intrinsic Dimension Estimation, to directly learn the dynamic structure from experimental data and distinguish between thermal and scar initial states. This method is robust to a limited sample size and experimental measurement errors and is suitable for actual experimental environments.

Through this method, researchers hope to be able to identify the initial states that can exhibit the scar phenomenon without prior knowledge, only using data that is easily accessible experimentally, thereby providing new tools and perspectives for further exploring the properties and applications of quantum many - body scar systems.