Variational Quantum Hypernetworks for Generalized Quantum Circuit Initialization

Junyong Lee, Jeihee Cho, Daniel Justice, Euimin Lee, Hyeonseong Jung, Shiho Kim

Keywords: Variational Quantum Algorithms, Barren Plateaus, Meta-Learning, Parameter Initialization, Quantum Machine Learning

TL;DR: We built a VQ-Hypernetwork that meta-learns how to find the depth-aware starting point for parameterized quantum circuit, alleviating the vanishing/exploding gradient problem to make training stable, fast, and effective.

Abstract:

To the Program Chairs,

We are withdrawing our submission (ID: 23591) to incorporate significant revisions based on the reviewers' valuable feedback.

Since we plan to resubmit this work to another double-blind venue shortly, we are concerned that leaving the abstract publicly visible on OpenReview may compromise the anonymity of our future submission. Additionally, given the recent community discussions regarding the protection of pre-publication ideas, we explicitly request that the abstract text be removed or hidden from the public view.

We believe this measure is necessary to ensure fair review in our subsequent submission. Thank you for your support.

Sincerely, The Authors

Primary Area: optimization

Submission Number: 23591