Krylov Delocalization/Localization across Ergodicity Breaking
Heiko Georg Menzler, Rishabh Jha

TL;DR
This paper explores how operator complexity growth relates to localization phenomena on the Krylov chain, revealing a transition between delocalization in ergodic regimes and localization in weakly ergodic regimes, with implications for understanding ergodicity breaking.
Contribution
It introduces a phenomenological framework linking ergodicity breaking to localization transitions on the Krylov chain, supported by analysis of the SYK and East models.
Findings
Delocalization occurs in the ergodic regime of the Krylov chain.
Localization occurs in weakly ergodic regimes.
Collapse of operator dynamics is observed in the ergodic regime.
Abstract
Krylov complexity has recently gained attention where the growth of operator complexity in time is measured in terms of the off-diagonal operator Lanczos coefficients. The operator Lanczos algorithm reduces the problem of complexity growth to a single-particle semi-infinite tight-binding chain (known as the Krylov chain). Employing the phenomenon of Anderson localization, we propose the phenomenology of inverse localization length on the Krylov chain that undergoes delocalization/localization transition on the Krylov chain while the physical system undergoes ergodicity breaking. On the Krylov chain we find delocalization in an ergodic regime, as we show for the SYK model, and localization in case of a weakly ergodicity-broken regime. Considering the dynamics beyond scrambling, we find a collapse across different operators in the ergodic regime. We test for two settings: (1) the coupled…
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.
Taxonomy
TopicsComplex Systems and Time Series Analysis · Neural Networks and Applications · Opinion Dynamics and Social Influence
