Quantum Wavefunction Augmentation via Variational Autoencoders

Abstract

Sample-based quantum diagonalization (SQD) has emerged as a promising route for quantum-centric supercomputing, relying on classical diagonalization of the molecular Hamiltonian within a hardware-sampled determinant subspace. However, its accuracy degrades in strongly correlated regimes where the relevant determinant space exceeds what finite-shot sampling can capture. In this work, we introduce Quantum Wavefunction Augmentation via Variational Autoencoders (Q-WAVE), a hybrid method that combines determinants sampled via SqDRIFT Krylov circuits and configuration interaction singles and doubles (CISD) determinants with generative machine learning. Using a custom β-annealed variational autoencoder (VAE) model, Q-WAVE iteratively expands this basis toward the variational ground state. The VAE learns the wavefunction's primary support structure from the combined hardware and CISD seed in a continuous latent space, generating new dominant determinants beyond any fixed excitation hierarchy. The resulting compact wavefunction exceeds what can be extracted from raw hardware samples alone. We demonstrate sub-millihartree accuracy compared to full configuration interaction for H₂O and N₂ dissociation. Finally, we establish Q-WAVE's scalability on a 52-qubit ethylene system (achieving sub-millihartree accuracy versus CCSD(T)) and a highly correlated 60-qubit Cr₂ stress test that attains chemical accuracy upon a final perturbative correction.

Learning Disentangled Representations with Quantum Variational Autoencoders
Gaoyuan Wang, Jerry Tan, Mark Gerstein
arXiv5 Oct 2026PreprintCC BY 4.0Full textQuantum physics & quantum informationarXiv
Band structure of two-dimensional graphene and of one-dimensional nanoribbons via variational quantum algorithms: from energies to wavefunction quality
Raúl Medrano Millán, Rodrigo Gil-Merino y Rubio, Juan Borge +1
arXiv3 Nov 2026PreprintCC BY 4.0Full textCondensed matter Quantum physics & quantum informationarXiv
Elucidating structure-property correlations in ferroelectric Hf0.5Zr0.5O2 films using variational autoencoders
Kévin Alhada-Lahbabi, Brice Gautier, Damien Deleruyelle +1
Materials & Design1 Oct 2026Peer-reviewedCC BY3 citedGeneral physicsHAL
Smart Machine Learning for Solving the Full Electron-Nuclear Schrödinger Equation with a Quantum-Chemically Inspired Wavefunction Ansatz
Yaolong Zhang, Bin Jiang, Hua Guo
arXiv30 Sep 2026PreprintarXiv licenceFull textPhysical chemistry Computational physicsarXiv
Emergent classicality and wavefunction branching in an isolated quantum many-body system
Saúl Pilatowsky-Cameo, Jordan Cotler, Daniel Ranard +1
arXiv16 Sep 2026PreprintarXiv licenceFull textMathematical physics Condensed matterarXiv

Cite

BibTeX RIS

BibTeX

@misc{halder2026quantum,
  title = {{Quantum Wavefunction Augmentation via Variational Autoencoders}},
  author = {Sonaldeep Halder and Chayan Patra and Rahul Maitra},
  howpublished = {arXiv},
  year = {2026},
  month = sep,
  doi = {10.48550/arxiv.2609.07972},
  eprint = {2609.07972},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2609.07972},
  note = {CC BY 4.0},
}

RIS

TY  - UNPB
TI  - Quantum Wavefunction Augmentation via Variational Autoencoders
AU  - Sonaldeep Halder
AU  - Chayan Patra
AU  - Rahul Maitra
PY  - 2026
DA  - 2026/09/07
PB  - arXiv
DO  - 10.48550/arxiv.2609.07972
UR  - https://arxiv.org/abs/2609.07972
AB  - Sample-based quantum diagonalization (SQD) has emerged as a promising route for quantum-centric supercomputing, relying on classical diagonalization of the molecular Hamiltonian within a hardware-sampled determinant subspace. However, its accuracy degrades in strongly correlated regimes where the relevant determinant space exceeds what finite-shot sampling can capture. In this work, we introduce Quantum Wavefunction Augmentation via Variational Autoencoders (Q-WAVE), a hybrid method that combines determinants sampled via SqDRIFT Krylov circuits and configuration interaction singles and doubles (CISD) determinants with generative machine learning. Using a custom β-annealed variational autoencoder (VAE) model, Q-WAVE iteratively expands this basis toward the variational ground state. The VAE learns the wavefunction's primary support structure from the combined hardware and CISD seed in a continuous latent space, generating new dominant determinants beyond any fixed excitation hierarchy. The resulting compact wavefunction exceeds what can be extracted from raw hardware samples alone. We demonstrate sub-millihartree accuracy compared to full configuration interaction for H_2O and N_2 dissociation. Finally, we establish Q-WAVE's scalability on a 52-qubit ethylene system (achieving sub-millihartree accuracy versus CCSD(T)) and a highly correlated 60-qubit Cr_2 stress test that attains chemical accuracy upon a final perturbative correction.
ER  -

Abstract

Related

Cite