Classical State Preparation for Variational Quantum Algorithms via Reinforcement Learning
The paper presents a new framework called CRiSP, which utilizes reinforcement learning for classical state preparation in variational quantum algorithms. This approach aims to address challenges such as barren plateaus and local minima that hinder optimization. Evaluations show that CRiSP significantly outperforms existing initialization methods in terms of energy accuracy on various benchmarks.
- ▪CRiSP formulates discrete prefix selection as a sequential decision-making problem using reinforcement learning.
- ▪The framework employs Neural-Guided Monte Carlo Tree Search and a Transformer-based policy for state preparation.
- ▪CRiSP has been evaluated on QAOA benchmarks and shows a mean improvement of 3.17 times in average energy accuracy compared to state-of-the-art methods.
Opening excerpt (first ~120 words) tap to expand
Quantum Physics arXiv:2605.23138 (quant-ph) [Submitted on 22 May 2026] Title:Classical State Preparation for Variational Quantum Algorithms via Reinforcement Learning Authors:Gino Kwun, Dhanvi Bharadwaj, Gokul Subramanian Ravi View a PDF of the paper titled Classical State Preparation for Variational Quantum Algorithms via Reinforcement Learning, by Gino Kwun and 2 other authors View PDF HTML (experimental) Abstract:Variational Quantum Algorithms (VQAs) potentially offer a pathway to practical quantum advantage, but their optimization is heavily hindered by barren plateaus and numerous local minima.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv cs.AI.
Discussion
0 commentsMore from arXiv cs.AI
