aiqbench analysis
This research proposes a qubit-efficient framework for continuous optimization on standard gate-based quantum computers by encoding variables into single-qubit amplitudes rather than binary representations. By leveraging amplitude estimation, the method replaces the logarithmic qubit cost of discretization with a constant one-qubit cost per variable, shifting precision requirements to the estimation procedure. This approach offers a distinct tradeoff between circuit width and solution precision, potentially enabling higher-resolution optimization without expanding qubit counts.
High Precision Qubit-Efficient Variational Continuous Optimization via Amplitude Estimation
arXiv:2606.08690v1 Announce Type: new Abstract: Optimization of continuous-variable objectives on standard gate-based quantum computers via variational algorithms such as QAOA is typically approached by first discretizing each decision variable into a finite binary representation. This increases qubit requirements and restricts solution precision through fixed-resolution encodings. We propose a qubit-efficient variational framework for continuous optimization that instead encodes each decision variable into the squared amplitude or equivalently, the measurement probability of a single qubit state. This removes explicit discretization from the variable representation while remaining entirely within the standard qubit circuit model unlike methods like CV-QAOA employing qumode based hardware to achieve the same. To read out encoded variables, we propose using amplitude estimation rather than naive sampling or tomographic reconstruction, with the goal of improving precision scaling for continuous-value recovery. We outline how amplitude-estimation error propagates to decision-variable error and then to objective-value error under standard regularity assumptions, suggesting a distinct width-versus-precision tradeoff relative to discretized approaches. In particular, the framework replaces the logarithmic increase in qubits needed for finer binary precision with a constant cost of one qubit per decision variable, while shifting accuracy requirements into the estimation procedure. We position this approach relative to traditional discretized variational formulations, and argue that it provides a promising new direction for continuous optimization on standard qubit architectures.
Abstract:Optimization of continuous-variable objectives on standard gate-based quantum computers via variational algorithms such as QAOA is typically approached by first discretizing each decision variable into a finite binary representation. This increases qubit requirements and restricts solution precision through fixed-resolution encodings. We propose a qubit-efficient variational framework for continuous optimization that instead encodes each decision variable into the squared amplitude or equivalently, the measurement probability of a single qubit state. This removes explicit discretization from the variable representation while remaining entirely within the standard qubit circuit model unlike methods like CV-QAOA employing qumode based hardware to achieve the same. To read out encoded variables, we propose using amplitude estimation rather than naive sampling or tomographic reconstruction, with the goal of improving precision scaling for continuous-value recovery. We outline how amplitude-estimation error propagates to decision-variable error and then to objective-value error under standard regularity assumptions, suggesting a distinct width-versus-precision tradeoff relative to discretized approaches. In particular, the framework replaces the logarithmic increase in qubits needed for finer binary precision with a constant cost of one qubit per decision variable, while shifting accuracy requirements into the estimation procedure. We position this approach relative to traditional discretized variational formulations, and argue that it provides a promising new direction for continuous optimization on standard qubit architectures.
arXiv:2606.08690 [quant-ph]
(or arXiv:2606.08690v1 [quant-ph] for this version)
https://doi.org/10.48550/arXiv.2606.08690 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
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 · 302 words