Research map / Probabilistic and scientific modeling / Molecules and proteins
Multimodal Crystal Flow: Any-to-Any Modality Generation for Unified Crystal Modeling
Kiyoung Seong, Sungsoo Ahn, Sehui Han, Changyoung Park
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 inter-orbit permutation augmentation (ablation) | MCFlow-B · MP-20 CSP | match rate (%): 60.41 -> 63.14 | Improved |
| add inter-orbit permutation augmentation (ablation) | MCFlow-B · MPTS-52 CSP | match rate (%): 22.80 -> 26.46 | Improved |
| add inter-orbit permutation augmentation (ablation) | MCFlow-B · MP-20 CSP | RMSE: 0.0772 -> 0.0611 | Improved |
| add intra-orbit permutation augmentation (ablation) | MCFlow-B · MP-20 CSP | match rate (%): 58.54 -> 60.41 | Improved |
| add intra-orbit permutation augmentation (ablation) | MCFlow-B · MPTS-52 CSP | match rate (%): 18.80 -> 22.80 | Improved |
| add positional embedding based on proposed atom ordering (ablation) | MCFlow-B · MP-20 CSP | match rate (%): 54.83 -> 58.54 | Improved |
| add positional embedding based on proposed atom ordering (ablation) | MCFlow-B · MPTS-52 CSP | match rate (%): 16.44 -> 18.80 | Improved |
| add positional embedding based on proposed atom ordering (ablation) | MCFlow-B · MP-20 DNG | compositional validity (%): 85.30 -> 89.30 | Improved |
| add positional embedding based on proposed atom ordering (ablation) | MCFlow-B · MPTS-52 DNG | compositional validity (%): 75.20 -> 84.83 | Improved |
| add inter/intra permutation augmentation (ablation) | MCFlow-B · MP-20 DNG | structural validity (%): 99.35 -> 99.44 | No clear effect |
| scale model from base to large, single sample | MCFlow · MPTS-52 CSP | RMSE: 0.1572 -> 0.1401 | Improved |
| scale model from base to large, 20 samples | MCFlow · MP-20 CSP | match rate (%): 77.12 -> 76.08 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- FlowMM (1) — FlowMM: Generating Materials with Riemannian Flow Matching (ICML 2024) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- LLM Meets Diffusion: A Hybrid Framework for Crystal Material Generation (NeurIPS 2025 · similarity 0.90)
- CrystalREPA: Transferring Physical Priors from Universal MLIPs to Crystal Generative Models (NeurIPS 2026 · similarity 0.90)
- EP-Flow: Disordered Crystal Structure Prediction without Site-level Annotation (NeurIPS 2026 · similarity 0.89)
- Scalable Diffusion for Materials Generation (ICLR 2024 · similarity 0.89)
- All-atom Diffusion Transformers: Unified generative modelling of molecules and materials (ICML 2025 · similarity 0.88)
- WyckoffDiff -- A Generative Diffusion Model for Crystal Symmetry (ICML 2025 · similarity 0.88)
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.
