Self-Refinement of Auxiliary-Field Quantum Monte Carlo via Non-Orthogonal Configuration Interaction
Zoran Sukurma, Martin Schlipf, Georg Kresse · 2025
Read Self-Refinement of Auxiliary-Field Quantum Monte Carlo via Non-Orthogonal Configuration Interaction by Zoran Sukurma, Martin Schlipf, Georg Kresse (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-12765.txt.
Description
Quantum Monte Carlo methods can compute molecular energies with superb accuracy, but one popular variant depends heavily on the quality of a trial wave function fed into it. Zoran Sukurma, Martin Schlipf, and Georg Kresse remove that bottleneck. Their algorithm lets the simulation choose its own trial ingredients: as the random walk proceeds, it keeps the electron configurations that contribute most to a refined energy estimate.
With 100 to 200 configurations, errors fell by up to a factor of ten for second-row elements that simpler trial states describe poorly. On a standard test set of small molecules, average errors dropped inside chemical accuracy. For benzene, the largest system tried, error fell 80 percent, though computing time grew tenfold. The authors note the method still struggles where electrons are strongly correlated or spin is contaminated.
Catalog record
| CREATOR | Zoran Sukurma, Martin Schlipf, Georg Kresse |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 43 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-12765