An Arrow of Time in Stable Autonomous Generation
Abstract
Does stable autonomous generation select a direction of time? This question matters both for learning models that autonomously generate prescribed dynamics and for designing controllers that make an invariant set attracting without changing its dynamics. Periodic motion and quasiperiodic torus rotations admit stable generation in either direction. For a chaotic attractor, however, reversing the equations reproduces its backward trajectories while turning attraction into repulsion. Can additional latent variables or autonomous feedback restore attraction without changing the reversed motion? We prove an obstruction to such exact generation by finite-dimensional autonomous systems with C2 vector fields, compact attracting sets, and continuous readouts. The result applies to the classical Lorenz attractor and, more generally, to dynamics with an ergodic invariant measure of positive metric entropy whose intrinsic local stable fibers are totally disconnected on a set of positive measure. The proof connects this geometry to an entropy contradiction: a hypothetical attracting realization would force the output entropy to vanish, contrary to the prescribed chaotic dynamics. We use reservoir-computing examples to illustrate the difficulty of stable autonomous generation for time-reversed chaos: both time directions are closely tracked while the networks are driven by data, but only the forward example retains Lorenz geometry during autonomous generation. The theorem identifies a class of dynamics that cannot be learned for exact, stable autonomous generation by finite-dimensional smooth models satisfying the stated assumptions, regardless of the learning procedure or the number of latent variables.
Related
- Single-photon time qubit: Quantum-vacuum control and complementarity of the arrow of time
- Entropy and Frenesy: the arrow and the bustle of time
- From Autonomous Knowledge Generation to Generalizable Property Prediction: A Precision-First Pipeline for Electro-Optic Materials Discovery
- Generation of Stable Peak-Power Similaritons through Gain-Managed Nonlinearity
- Stable tracking of turbulence-disturbed laser communication spots using a time-series state fusion network
BibTeX
@misc{lu2026arrow,
title = {{An Arrow of Time in Stable Autonomous Generation}},
author = {Zhixin Lu},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.02845},
eprint = {2610.02845},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.02845},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - An Arrow of Time in Stable Autonomous Generation
AU - Zhixin Lu
PY - 2026
DA - 2026/10/02
PB - arXiv
DO - 10.48550/arxiv.2610.02845
UR - https://arxiv.org/abs/2610.02845
AB - Does stable autonomous generation select a direction of time? This question matters both for learning models that autonomously generate prescribed dynamics and for designing controllers that make an invariant set attracting without changing its dynamics. Periodic motion and quasiperiodic torus rotations admit stable generation in either direction. For a chaotic attractor, however, reversing the equations reproduces its backward trajectories while turning attraction into repulsion. Can additional latent variables or autonomous feedback restore attraction without changing the reversed motion? We prove an obstruction to such exact generation by finite-dimensional autonomous systems with C^2 vector fields, compact attracting sets, and continuous readouts. The result applies to the classical Lorenz attractor and, more generally, to dynamics with an ergodic invariant measure of positive metric entropy whose intrinsic local stable fibers are totally disconnected on a set of positive measure. The proof connects this geometry to an entropy contradiction: a hypothetical attracting realization would force the output entropy to vanish, contrary to the prescribed chaotic dynamics. We use reservoir-computing examples to illustrate the difficulty of stable autonomous generation for time-reversed chaos: both time directions are closely tracked while the networks are driven by data, but only the forward example retains Lorenz geometry during autonomous generation. The theorem identifies a class of dynamics that cannot be learned for exact, stable autonomous generation by finite-dimensional smooth models satisfying the stated assumptions, regardless of the learning procedure or the number of latent variables.
ER -