aiqbench analysis
This research introduces a conditioning mechanism into generative quantum denoising diffusion models, allowing the generation of states from multiple target distributions while sharing parameters to improve efficiency. Numerical simulations demonstrate that this approach reduces state generation error by over an order of magnitude across tasks ranging from single-qubit preparation to many-body ground state generation. The work represents a significant methodological advance in generative quantum machine learning by addressing the practical challenge of training separate models for distinct distributions.
Conditioning in Generative Quantum Denoising Diffusion Models
arXiv:2509.17569v2 Announce Type: replace Abstract: Quantum denoising diffusion models have recently emerged as a powerful framework for generative quantum machine learning. In this work, we extend these models by introducing a conditioning mechanism that enables the generation of quantum states drawn from multiple target distributions. By sharing parameters across distinct classes of quantum states, our approach avoids the need to train separate models for each distribution. We validate our method through numerical simulations that span single-qubit generation tasks, entangled state preparation, and many-body ground state generation. Across these tasks, conditioning significantly reduced the error of targeted state generation by more than an order of magnitude. Finally, we perform an ablation study to quantify the effect of key hyperparameters on the model performance.
Abstract:Quantum denoising diffusion models have recently emerged as a powerful framework for generative quantum machine learning. In this work, we extend these models by introducing a conditioning mechanism that enables the generation of quantum states drawn from multiple target distributions. By sharing parameters across distinct classes of quantum states, our approach avoids the need to train separate models for each distribution. We validate our method through numerical simulations that span single-qubit generation tasks, entangled state preparation, and many-body ground state generation. Across these tasks, conditioning significantly reduced the error of targeted state generation by more than an order of magnitude. Finally, we perform an ablation study to quantify the effect of key hyperparameters on the model performance.
Close to published version. 17 pages, 13 figures, 3 tables. Code, notebooks, and data at this https URL
arXiv:2509.17569 [quant-ph]
(or arXiv:2509.17569v2 [quant-ph] for this version)
https://doi.org/10.48550/arXiv.2509.17569 Focus to learn more arXiv-issued DOI via DataCite
Quantum Sci. Technol. 11 (3), 035047 (2026)
https://doi.org/10.1088/2058-9565/ae8881 Focus to learn more DOI(s) linking to related resources
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 241 words
Related coverage
Researchers Derive Parity Floor for Quantum Denoisers
Quantum Zeitgeist · Aug 17, 2026
Discrete Diffusion with Sample-Efficient Estimators for Conditionals
arXiv · Jul 22, 2026
Thirty minutes to better weather insights with Earth2Studio
Quantum Zeitgeist · Sep 22, 2026
Inception in DiffusionGemma - Jailbreaking a Diffusion Language Model by Pinning Tokens Anywhere on the Canvas
LessWrong · Jul 23, 2026