Research map / Graph learning / GNN architectures
From atom to space: A region-based readout function for spatial properties of materials
Jiawen Zou, Weimin Tan, Zhongyao Wang, Hao Qi, Bo Yan
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add spatial nodes to PaiNN (pool spatial nodes) (ablation) | PaiNN · MOF C3H6/C3H8 separation | R2: 0.691 -> 0.794 | Improved |
| add spatial nodes to PaiNN (ablation) | PaiNN · COF CH4 adsorption | R2: 0.736 -> 0.979 | Improved |
| add spatial nodes to PaiNN (ablation) | PaiNN · zeolite CH4 heat | R2: 0.791 -> 0.886 | Improved |
| add multimodal transformer on top of spatial nodes (ablation) | PaiNN + SN · spatial properties (Table 1) | R2: 0.794 -> 0.784 (MOF C3H6/C3H8), 0.886 -> 0.969 (zeolite) | No clear effect |
| add GraphTrans readout to PaiNN | PaiNN · MOF N2 adsorption | R2: 0.925 -> 0.912 | No clear effect |
| add SpatialRead (SN+MM) to pretrained JMP | JMP · MOF N2 adsorption | R2: 0.971 -> 0.988 | Improved |
| add SpatialRead (SN+MM) to JMP | JMP · COF CH4 adsorption | R2: 0.884 -> 0.982 | Improved |
| add spatial nodes only to PaiNN (ablation) | PaiNN · ASA geometric property | R2: 0.993 -> 0.974 | Hurt |
| add multimodal transformer to PaiNN + SN (ablation) | PaiNN + SN · PLD | R2: 0.856 -> 0.965 | Improved |
| add multimodal transformer to PaiNN + SN (ablation) | PaiNN + SN · LCD | R2: 0.913 -> 0.975 | Improved |
| add SpatialRead to JMP | JMP · MatBench MP Formation Energy | MAE: 13.6 -> 15.3 | Hurt |
| add SpatialRead to JMP | JMP · MatBench MP Gap | MAE: 0.119 -> 0.107 | Improved |
Closest papers by meaning
- Learning the Electronic Hamiltonian of Large Atomic Structures (ICML 2025 · similarity 0.88)
- AtomWorld: A Benchmark for Evaluating Spatial Reasoning in Large Language Models on Material Structures (ICML 2026 · similarity 0.87)
- Vector Field Oriented Diffusion Model for Crystal Material Generation (AAAI 2024 · similarity 0.85)
- Crystalformer: Infinitely Connected Attention for Periodic Structure Encoding (ICLR 2024 · similarity 0.85)
- Offline Materials Optimization with CliqueFlowmer (NeurIPS 2026 · similarity 0.84)
- Learning Condensed Graph via Differentiable Atom Mapping for Reaction Yield Prediction (ICML 2025 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
