Gradient-estimator design overcomes trainability barriers in neural-network-based variational optimization
Neural networks provide expressive representations for scientific computing. However, even sufficiently expressive networks can suffer training failure in weak-gradient regimes, limiting their practical use in quantum many-body physics and ab initio quantum chemistry. Here we derive an unbiased direct gradient estimator and introduce the adaptive minimum-variance phase (AMVP) estimator for neural-network variational optimization. By improving the signal-to-noise ratio of weak gradients, these methods enable reliable scientific calculations where training previously failed, while substantially reducing computational cost. The framework enables compact networks to outperform larger and fine-tuned default standard-estimator models with over an order of magnitude less GPU time on correlated flux models, and ultimately exceed the density matrix renormalization group (DMRG) accuracy. It further achieves chemical accuracy in N2 bond breaking and, for the first time, in heavy-element I2 with explicit spin–orbit coupling. These results demonstrate that gradient-estimator design expands the capabilities of neural-network variational methods for accurate scientific computing.
Authors
- Yi-Ran Xue (ORCID: https://orcid.org/0009-0005-0649-7323)
- Chenan Wei (ORCID: https://orcid.org/0000-0002-1263-7354)
- Rui Wang
- Baigeng Wang
Institutions
- University of Massachusetts Amherst (US)
- Nanjing University (CN)
Publication Details
- Journal
- Research Square
- Published
- 2026-11-19
- DOI
- https://doi.org/10.21203/rs.3.rs-11582001/v1
- Primary Topic
- Quantum Computing Algorithms and Architecture
- Type
- preprint
Funders
- National Natural Science Foundation of China
- Natural Science Foundation of Jiangsu Province
- Fundamental Research Funds for the Central Universities