Temporal Coarse-Graining as the Origin of Macroscopic Friction in Quantum Spin Chains via Data-Driven Liouvillian Extraction
Seiki Saito

TL;DR
This paper presents a data-driven approach to derive macroscopic friction and viscosity in quantum spin chains, revealing their dependence on the observer's temporal resolution and the necessity of finite coarse-graining.
Contribution
It introduces a generalized EDMD framework with Mori-Zwanzig projection to extract hydrodynamic coefficients directly from quantum dynamics, highlighting the role of temporal coarse-graining.
Findings
Mechanical elasticity is derived from unitary dynamics, preserving reversibility.
Macroscopic friction and viscosity oscillate around zero, indicating no net dissipation at exact limits.
Finite observation timescales induce a crossover regime with positive friction and viscosity.
Abstract
Understanding the emergence of macroscopic irreversible hydrodynamics from the reversible unitary dynamics of isolated quantum many-body systems remains a fundamental challenge. Conventional approaches often force spin density dynamics into purely diffusive models, obscuring the microscopic interplay of pressure, spin current, and local friction. Furthermore, reconciling true irreversibility with strictly unitary evolution raises profound questions about the role of the observer's temporal resolution. In this paper, we introduce a fully data-driven framework based on generalized Extended Dynamic Mode Decomposition (gEDMD) integrated with the Mori-Zwanzig projection. By expanding the observable dictionary to explicitly include spin currents, we directly extract the Navier-Stokes hydrodynamic coefficients from a chaotic XXZ spin chain across varying temporal coarse-graining scales. Our…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
