mlsanityData to 14 Oct 2026

CrystalFormer-CSP: Thinking Fast and Slow for Crystal Structure Prediction

arXiv 2512.18251 ↗First posted 20 Dec 2025cond-mat.mtrl-sci cs.LG physics.comp-phcited by 1

Not on the map. Why?

Abstract (arXiv)

Crystal structure prediction is a fundamental problem in materials science. We present CrystalFormer-CSP, an efficient framework that unifies data-driven heuristic and physics-driven optimization approaches to predict stable crystal structures for given chemical compositions. The approach combines pretrained generative models for space-group-informed structure generation and a universal machine learning force field for energy minimization. Reinforcement fine-tuning can be employed to further boost the accuracy of the framework. We demonstrate the effectiveness of CrystalFormer-CSP on benchmark problems and showcase its usage via web interface and language model integration.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 15 of its 48 references that are in the map. First posted on arXiv 20 Dec 2025.

Builds on

15 of 48 references.

References · 15

  1. 2022Training language models to follow instructions with human feedbackcited by 15,971
  2. 2017Proximal Policy Optimization Algorithmscited by 15,215
  3. 2022Classifier-Free Diffusion Guidancecited by 5,847
  4. 2021Crystal Diffusion Variational Autoencoder for Periodic Material Generationcited by 270
  5. 2023Language Modeling Is Compressioncited by 225
  6. 2025RL's Razor: Why Online Reinforcement Learning Forgets Lesscited by 176
  7. 2023Crystal Structure Prediction by Joint Equivariant Diffusioncited by 148
  8. 2024FlowMM: Generating Materials with Riemannian Flow Matchingcited by 109
  9. 2024Space Group Informed Transformer for Crystalline Materials Generationcited by 41
  10. 2025Reinforcement Fine-Tuning for Materials Designcited by 13
  11. 2025Equivariant Diffusion for Crystal Structure Predictioncited by 9
  12. 2025AI-Driven Expansion and Application of the Alexandria Databasecited by 5
3 more
  1. 2023Shotgun crystal structure prediction using machine-learned formation energiescited by 4
  2. 2025Self-Optimizing Machine Learning Potential Assisted Automated Workflow for Highly Efficient Complex Systems Material Designcited by 1
  3. 2025High-dimensional Optimization with Low Rank Tensor Sampling and Local Searchcited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Multimodal Crystal Flow: Any-to-Any Modality Generation for Unified Crystal Modelingcited by 7

Similar, not cited

Close in wording, no citation either way

  1. 2026uFlowCSP: Crystal Structure Prediction using Mean flow generative models
  2. 2023Data-Driven Score-Based Models for Generating Stable Structures with Adaptive Crystal Cells
  3. 2025Introducing physics-informed generative models for targeting structural novelty in the exploration of chemical space
  4. 2026CrystalReasoner: Reasoning and RL for Property-Conditioned Crystal Structure Generation
  5. 2026MatMind: A Structure-Activity Knowledge-Driven Generative Foundation Model for Materials Science
  6. 2026Riemannian Flow Models with Reinforcement Learning for Molecular Crystal Structure Prediction
  7. 2026Composable Crystals: Controllable Materials Discovery via Concept Learning
  8. 2024Unleashing the power of novel conditional generative approaches for new materials discovery