DisQu: Investigating the Impact of Disorder in Quantum Generative Models

Disordered Quantum many-body Systems (DQS) and Quantum Neural Networks (QNN) have many structural features in common. However, a DQS is essentially an initialized QNN with random weights, often leading to non-random outcomes. In this work, we emphasize the possibilities of random processes being a deceptive quantum-generating model effectively hidden in a QNN. When we choose weights in a QNN randomly the unitarity property of quantum gates is unchanged. As we show, this can lead to memory effects with multiple consequences on the learnability and trainability of QNN one would not expect from a classical neural network with random weights. This phenomenon may lead to a fundamental misunderstanding of the capabilities of common quantum generative models, where the generation of new samples is essentially averaging over random outputs. While we suggest that DQS can be effectively used for tasks like image augmentation, we draw the attention that overly simple datasets are often used to show the generative capabilities of quantum models, potentially leading to overestimation of their effectiveness.

💡 Research Summary

The paper “DisQu: Investigating the Impact of Disorder in Quantum Generative Models” examines how random‑weight quantum neural networks (QNNs) are essentially disordered quantum many‑body systems (DQSs) and how this hidden disorder can masquerade as generative capability. The authors begin by noting that a QNN consists of an encoding stage, a trainable unitary layer, and a measurement stage. When the trainable parameters are drawn randomly, the resulting unitary is mathematically equivalent to the time‑evolution operator of a DQS with a random Hamiltonian. Crucially, the unitarity of quantum gates is preserved, so the system remains a legitimate quantum process despite the randomness.

The study focuses on a one‑dimensional XXZ‑type Hamiltonian
(\hat H = J\sum_j (\hat c^\dagger_j \hat c_{j+1}+h.c.) + V\sum_j \hat n_j \hat n_{j+1} + \sum_j d_j \hat n_j),
where the disorder potentials (d_j) are drawn from a uniform box distribution (