An accurate Lanczos method for the matrix product state representation

Abstract

The Lanczos algorithm naturally finds multiple low-lying eigenstates simultaneously, providing a global and robust treatment of the spectrum without local minima or root flipping. However, when implemented within the matrix product state formalism, its accuracy is severely affected by bond truncation, leading to a pronounced high-error plateau. Here, we show the modified thick-block Lanczos method to enhance the convergence of the Lanczos algorithm within the matrix product state formalism. Numerical benchmarks on the Fermi-Hubbard and Heisenberg models demonstrate that our method achieves the optimal accuracy allowed by the bond dimension. In addition, we benchmark it on a 50-site Fermi-Hubbard model and a 120-site Heisenberg model to demonstrate its scalability. Furthermore, when incorporating the shift-and-invert technique, our method enables the direct targeting of multiple excited states without finding lower eigenstates.

Evaluation of 2νββ decay nuclear matrix elements using Lanczos techniques preserving Gamow-Teller strengths from both parent and daughter nuclei
Deepak Patel, Anil Kumar, Noritaka Shimizu +1
arXiv2 Nov 2026PreprintCC BY 4.0Full textNuclear physicsarXiv
Accurate representation of the continuum orbital in a Gaussian basis set: application to Auger decay rates
Dibyendu Mahato, Saikat Roy, Wojciech Skomorowski
arXiv3 Nov 2026PreprintCC BY 4.0Full textPhysical chemistry Atomic, molecular & optical physicsarXiv
Efficient time-dependent dynamics of matrix product states using average Hamiltonians
Belal Abouraya, Jirawat Saiphet, Fedor Jelezko +1
Physical Review Research6 Oct 2026Peer-reviewedCC BY 4.0Atomic, molecular & optical physicsCrossref
A New Perspective on Matrix Representation of Paraxial Geometric Optics using Two Kinds of Three-Matrix Decompositions of the 2× 2 Special-Linear-Group Matrices
Satoshi Itoh
arXiv17 Oct 2026PreprintCC BY-NC-SA 4.0Full textOptics & photonicsarXiv
A first introduction to Matrix Product State algorithms for the integration of Lindblad equation
Christophe Chatelain
HAL15 Sep 2026PreprintCC BY-NC-ND 4.0Full textCondensed matter Quantum physics & quantum informationHALarXiv

Cite

BibTeX RIS

BibTeX

@article{wang2026accurate,
  title = {{An accurate Lanczos method for the matrix product state representation}},
  author = {Yu Wang and Zhangyu Yang and Xingyao Wu and Christian B. Mendl},
  journal = {Communications Physics},
  year = {2026},
  month = oct,
  doi = {10.1038/s42005-026-02896-x},
  url = {https://www.nature.com/articles/s42005-026-02896-x},
  note = {CC BY 4.0},
}

RIS

TY  - JOUR
TI  - An accurate Lanczos method for the matrix product state representation
AU  - Yu Wang
AU  - Zhangyu Yang
AU  - Xingyao Wu
AU  - Christian B. Mendl
PY  - 2026
DA  - 2026/10/14
JO  - Communications Physics
DO  - 10.1038/s42005-026-02896-x
UR  - https://www.nature.com/articles/s42005-026-02896-x
AB  - The Lanczos algorithm naturally finds multiple low-lying eigenstates simultaneously, providing a global and robust treatment of the spectrum without local minima or root flipping. However, when implemented within the matrix product state formalism, its accuracy is severely affected by bond truncation, leading to a pronounced high-error plateau. Here, we show the modified thick-block Lanczos method to enhance the convergence of the Lanczos algorithm within the matrix product state formalism. Numerical benchmarks on the Fermi-Hubbard and Heisenberg models demonstrate that our method achieves the optimal accuracy allowed by the bond dimension. In addition, we benchmark it on a 50-site Fermi-Hubbard model and a 120-site Heisenberg model to demonstrate its scalability. Furthermore, when incorporating the shift-and-invert technique, our method enables the direct targeting of multiple excited states without finding lower eigenstates.
SN  - 2399-3650
ER  -

Abstract

Related

Cite