Skip to content
DraconiticResearch
Search Sources API
203,203 studies31 sourcesUpdated 52 min ago
PreprintCC BY-NC-ND 4.0arXiv

Compressed magnetic Moment Tensor Potentials via low-rank matrix and tensor factorizations

  • Alexey S. Kotykhov
  • Alexander V. Shapeev
  • Maxim Rakhuba
  • Ivan S. Novikov
Open at sourcePDFDOICite

Abstract

We propose a parameter-reduced version of magnetic Moment Tensor Potential (mMTP) based on various matrix and tensor decompositions. The resulting compressed magnetic machine-learning potential reduces the number of parameters by a factor of approximately 1.5-3 without compromising predictive performance on the validation set. We evaluate the performance of the compressed potentials for magnetic, structural, and vibrational properties, as well as in molecular dynamics simulations of Fe-Al and CrN systems. We demonstrate that the simulation results obtained with the compressed mMTP are numerically consistent with those of the original uncompressed model and are in agreement with density functional theory calculations and experimental data.

Related

  1. Active learning of collinear magnetic Moment Tensor Potentials using the spin-MLIP package from soft-constrained spin-polarized DFT calculations: a case study of Fe-Pd
    Arseniy Burov, Alexey S. Kotykhov, Dmitry A. Aksyonov +2
    arXivPreprintCC BY-NC-ND 4.0Full textAtomic, molecular & optical physicsCondensed matterarXiv
  2. An Implicit, Moment-Conserving Low-Rank Tensor Method for the Hybrid Kinetic-Ion Fluid-Electron Plasma Model
    Stephen R. White, Adam J. Stanier, Luis Chacón +2
    arXivPreprintCC BY 4.0Full textPlasma physicsarXiv
  3. Convex optimization on moment polytopes: Hadamard mirror descent and efficient algorithms for quantum functionals and other tensor parameters
    Mahmut Levent Doğan, Keiya Sakabe, Michael Walter
    arXivPreprintCC BY 4.0Full textMathematical physicsOptimisation & controlarXiv
  4. Finite stress-tensor moment tomography of boundary-graviton coherences
    Dyuman Bhattacharya, Robert B. Mann
    arXivPreprintarXiv licenceFull textHigh-energy & particle physicsarXiv
  5. Inverse design of scalar-tensor EFT from odd-parity effective potentials
    Kazufumi Takahashi, Yuki Watanabe
    arXivPreprintCC BY 4.0Full textGravitation & general relativityHigh-energy & particle physicsarXiv

Cite

BibTeXRIS
BibTeX
@misc{kotykhov2026compressed,
  title = {{Compressed magnetic Moment Tensor Potentials via low-rank matrix and tensor factorizations}},
  author = {Alexey S. Kotykhov and Alexander V. Shapeev and Maxim Rakhuba and Ivan S. Novikov},
  howpublished = {arXiv},
  year = {2026},
  month = oct,
  doi = {10.48550/arxiv.2610.06158},
  eprint = {2610.06158},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2610.06158},
  note = {CC BY-NC-ND 4.0},
}
RIS
TY  - UNPB
TI  - Compressed magnetic Moment Tensor Potentials via low-rank matrix and tensor factorizations
AU  - Alexey S. Kotykhov
AU  - Alexander V. Shapeev
AU  - Maxim Rakhuba
AU  - Ivan S. Novikov
PY  - 2026
DA  - 2026/10/05
PB  - arXiv
DO  - 10.48550/arxiv.2610.06158
UR  - https://arxiv.org/abs/2610.06158
AB  - We propose a parameter-reduced version of magnetic Moment Tensor Potential (mMTP) based on various matrix and tensor decompositions. The resulting compressed magnetic machine-learning potential reduces the number of parameters by a factor of approximately 1.5-3 without compromising predictive performance on the validation set. We evaluate the performance of the compressed potentials for magnetic, structural, and vibrational properties, as well as in molecular dynamics simulations of Fe-Al and CrN systems. We demonstrate that the simulation results obtained with the compressed mMTP are numerically consistent with those of the original uncompressed model and are in agreement with density functional theory calculations and experimental data.
ER  -