# Scaling Law of Sim2Real transfer learning in expanding computational materials databases for real-world predictions

https://mdr.nims.go.jp/datasets/26be1168-7e03-4dc7-85b2-84ce7d70c18e

## File

- [s41524-025-01606-5.pdf](https://mdr.nims.go.jp/filesets/26eb6b00-c2db-41df-98c3-a23774d4c3ad/download) ([Detail](https://mdr.nims.go.jp/filesets/26eb6b00-c2db-41df-98c3-a23774d4c3ad.md))

## Id

26be1168-7e03-4dc7-85b2-84ce7d70c18e

## Local identifier



## Visibility

open_to_public

## State

published

## Created at

2025-06-13T02:43:45.875669Z

## Updated at

2025-06-13T07:30:25.767161Z

## Published at

2025-06-13T07:20:59.775500Z

## Doi



## First published url

https://doi.org/10.1038/s41524-025-01606-5

## Date published

2025-05-24

## Recorded date published



## Resource type

journal_article

## Manuscript type

vor

## Collection



## Title

- title: Scaling Law of Sim2Real transfer learning in expanding computational materials
    databases for real-world predictions
  title_type: original
  lang: en

## Description

- description: In this study, we demonstrate that the scaling law of simulation-to-real
    (Sim2Real) transfer learning holds for several machine learning tasks in materials
    science. Through three different prediction tasks for polymers and inorganic material
    systems, it was observed that the prediction error on real systems decreased monotonically
    with increasing the size of computational data according to a power law. Observing
    the scaling behavior offers various insights for advancing database development,
    such as determining the sample size necessary to achieve a desired predictive
    performance and a quantitative measure of the database’s potential value in real-world
    applications. Additionally, it aids in identifying equivalent sample sizes for
    physical and computational experiments and guiding the design of data production
    protocols principled the Sim2Realtransferability and scalability to downstream
    real-world tasks.
  description_type: abstract
  lang: und

## Creator

- name: Shunya Minami
  role: author
  orcid: https://orcid.org/0000-0002-3566-817X
- name: Yoshihiro Hayashi
  role: author
  orcid: https://orcid.org/0000-0002-7650-4083
- name: Stephen Wu
  role: author
  orcid: https://orcid.org/0000-0002-7847-8106
- name: Kenji Fukumizu
  role: author
- name: Hiroki Sugisawa
  role: author
- name: Masashi Ishii
  role: author
  orcid: https://orcid.org/0000-0003-0357-2832
- name: Isao Kuwajima
  role: author
  orcid: https://orcid.org/0000-0002-5994-3834
- name: Kazuya Shiratori
  role: author
- name: Ryo Yoshida
  role: author
  orcid: https://orcid.org/0000-0001-8092-0162

## Contact agent



## Publisher

organization: Springer Science and Business Media LLC

## Managing organization



## Keyword

- subject: molecular dynamics simulation
  schema: not_defined
- subject: scaling law of simulation- to-real
  schema: not_defined
- subject: Sim2Real
  schema: not_defined
- subject: transfer learning
  schema: not_defined
- subject: PoLyInfo
  schema: not_defined

## Rights

- identifier: https://creativecommons.org/licenses/by/4.0/

## Other identifier(s)



## Data origin

- data_origin_type: other

## Embargo



## Journal

- title: npj Computational Materials
  issn: '20573960'
  volume: '11'
  issue: '146'

## Conference



## Related item



## Funding

- identifier: hp210264
  funder_name: Ministry of Education, Culture, Sports, Science and Technology
- identifier: JPMJCR19I3, JPMJCR22O3, JPMJCR2332
  funder_name: MEXT | JST | Core Research for Evolutional Science and Technology
- identifier: 19H05820, 19H01132
  funder_name: MEXT | Japan Society for the Promotion of Science
- identifier: 23K19980
  funder_name: MEXT | Japan Society for the Promotion of Science
- identifier: 22K11949
  funder_name: MEXT | Japan Society for the Promotion of Science

## Instrument



## Instrument operator



## Instrument managing organization



## Measurement method



## Specimen



## Chemical composition



## Structure for specimen



## Structural feature for specimen



## Specific property for specimen



## Process for specimen treatment



## Computational method



## Energy level/transition state



## Software



## Custom property



## Fileset

- id: 26eb6b00-c2db-41df-98c3-a23774d4c3ad
  filename: s41524-025-01606-5.pdf
  content_type: application/pdf
  size: 4346757
  md5: d125c20ea29feb7564de24a3f4af733f

## Thumbnail

fileset_id: 26eb6b00-c2db-41df-98c3-a23774d4c3ad
filename: s41524-025-01606-5.pdf