DISTAL: Distillation and Self-Supervised Pretraining for Structure-Agnostic Materials Property Prediction
Not specified in the provided content
Abstract
DISTAL proposes a dual-prior framework for structure-agnostic materials property prediction that combines self-supervised compositional pretraining with structure-aware knowledge distillation.
Reality Card
DISTAL achieves the strongest overall performance among all evaluated feature combinations, improving over the reference benchmark on 37 out of 39 benchmark tasks.
The best-performing multimodal configuration combines compositional descriptors, pretrained latent features, and distilled structural features.
The main limitation is the reliance on a pretrained ALIGNN teacher, which may affect reproducibility if the teacher model is not accessible.
Paper to code
Verified implementation resources so builders can test the paper’s claims instead of stopping at the abstract.