Adaptive Dissipative State Preparation through Reinforcement Learning
Abstract
Dissipative algorithms approach the problem of ground state preparation by mimicking the natural thermalization of a quantum system in contact with a large, low-temperature thermal environment. The environment can be efficiently simulated by a single ancilla qubit with a variable energy gap that is repeatedly coupled to the system qubits to generate a dissipative channel, and then reset after each interaction. Here we present an adaptive implementation of the dissipative algorithm, RL-Adapt, that uses single-shot reinforcement learning to optimize the selection of ancilla frequency and system-bath interaction operator to maximize energy dissipation without relying on a priori knowledge of the system spectrum. The adaptive implementation results in significantly reduced convergence times and can successfully find the ground state even for non-ideal operator pools that fail to converge using non-adaptive, uniform random operator and ancilla frequency selection.
Related
- Classical State Preparation for Variational Quantum Algorithms via Reinforcement Learning
- Driven-Dissipative Ground State Preparation: Mixing Time and Randomness
- Locality-Preparation Time Tradeoffs in Dissipative Quantum Systems
- Locality in Dissipative Ground-State Preparation
- Boundary-rank obstructions and measurement-assisted recovery in dissipative flat-band preparation
BibTeX
@misc{myers2026adaptive,
title = {{Adaptive Dissipative State Preparation through Reinforcement Learning}},
author = {Nathan M. Myers and Chenxu Liu and Yulong Dong and Nicholas P. Bauman and Karol Kowalski},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.31370},
eprint = {2609.31370},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.31370},
}
RIS
TY - UNPB
TI - Adaptive Dissipative State Preparation through Reinforcement Learning
AU - Nathan M. Myers
AU - Chenxu Liu
AU - Yulong Dong
AU - Nicholas P. Bauman
AU - Karol Kowalski
PY - 2026
DA - 2026/09/25
PB - arXiv
DO - 10.48550/arxiv.2609.31370
UR - https://arxiv.org/abs/2609.31370
AB - Dissipative algorithms approach the problem of ground state preparation by mimicking the natural thermalization of a quantum system in contact with a large, low-temperature thermal environment. The environment can be efficiently simulated by a single ancilla qubit with a variable energy gap that is repeatedly coupled to the system qubits to generate a dissipative channel, and then reset after each interaction. Here we present an adaptive implementation of the dissipative algorithm, RL-Adapt, that uses single-shot reinforcement learning to optimize the selection of ancilla frequency and system-bath interaction operator to maximize energy dissipation without relying on a priori knowledge of the system spectrum. The adaptive implementation results in significantly reduced convergence times and can successfully find the ground state even for non-ideal operator pools that fail to converge using non-adaptive, uniform random operator and ancilla frequency selection.
ER -