Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
February 14, 2026Nature CommunicationsOpen Access

A data-efficient foundation model for porous materials based on expert-guided supervised learning

View Full Paper
Ask AI
Bookmark
Share

Authors

JZJiawen ZouZLZirui LvWTWeimin Tan

Discussion

Loading...

Member takes

Overview

Demonstrates how expert knowledge reduces data needs for modeling porous materials, suggesting a new approach in materials science.

Key Points

  • The aim is to develop a foundation model for porous materials using expert-guided supervised learning to limit data requirements.
  • Utilized expert knowledge to supervise pre-training of the model.
  • Developed potential energy surface basis functions for guest-host interactions.
  • Employed a multi-modal architecture to integrate data from material structure and energy surfaces.
  • Learned geometric features across different spatial scales for pre-training.
  • SpbNet outperformed models using significantly larger datasets, reducing errors by over 20%.
  • Demonstrated strong generalization across various materials, including Metal Organic Frameworks and zeolites.
  • Validated on over 50 downstream tasks related to adsorption and separation.

Cite This Study

Zou et al. (2026) studied this question.

synapsesocial.com/papers/699010382ccff479cfe56cebhttps://doi.org/10.1038/s41467-026-69245-y
View Full Paper
Ask AI
Bookmark
Share