Model Card for CrystaLLM-pi_mp_20_base

Model Details

Model Description

CrystaLLM-pi_mp_20_base is an unconditional generative model for crystal structures. It is a GPT-2 decoder-only model that generates Crystallographic Information Files (CIFs) from the patterns learned during training, with no property or diffraction conditioning attached.

The model was trained from scratch on MP-20 CIF text as an unconditional baseline for LeMat-Bench. For fine-tuning, use c-bone/CrystaLLM-pi_ft_alex_mp_20-text, which was initialised from the LeMat-Bulk pretrained model. Training used only the CIF text from c-bone/mp_20_pxrd, without the XRD conditioning vectors.

• Developed by: Bone et al. (University College London)
• Model type: Autoregressive Transformer, unconditional (~25.9M parameters)
• Language(s): CIF (Crystallographic Information File) syntax
• License: MIT
• Finetuned from model: None, trained from scratch

Model Sources

• Repository: GitHub: CrystaLLM-pi (generation)
• Dataset: HuggingFace: c-bone/mp_20_pxrd

Uses

Direct Use

Unconditional or prompt-steered generation of inorganic crystal structures, and evaluation as the MP-20 baseline in LeMat-Bench.

Out-of-Scope Use

• Property targeting: the model has no conditioning channel. Use a conditional model to request a property value.
• Disordered Systems: No native handling of partial occupancies or significant disorder.
• Organic/MOFs: Training data is inorganic, so organic frameworks are out of distribution.
• Extremely Large Unit Cells: the context window is 1024 tokens, so large cells will not fit.

Bias, Risks, and Limitations

• Distribution matching: output closely reflects the MP-20 distribution, so novelty is limited by design. Raise the generation temperature to move away from it.
• Validity is not stability: generated CIFs passing structural checks are not necessarily thermodynamically stable. Screen with an energy model before drawing conclusions.
• Computed data: training structures are relaxed DFT results rather than experimentally determined ones.

Getting started

For generation, use T2_load_and_generate.ipynb in CrystaLLM-pi. The training config is mp-20-text.jsonc.

Citation

@misc{bone2025discoveryrecoverycrystallinematerials,
      title={Discovery and recovery of crystalline materials with property-conditioned transformers},
      author={Cyprien Bone and Matthew Walker and Bradley A. A. Martin and Kuangdai Leng and Luis M. Antunes and Ricardo Grau-Crespo and Amil Aligayev and Javier Dominguez and Keith T. Butler},
      year={2025},
      eprint={2511.21299},
      archivePrefix={arXiv},
      primaryClass={cond-mat.mtrl-sci},
      url={https://arxiv.org/abs/2511.21299},
}