Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces UNATE, a framework for learning atomic embeddings from unlabeled crystal structures using a denoising autoencoder and contrastive learning. These embeddings are then used as input features for downstream crystal property prediction tasks. Results show a 2.7% improvement over a full-data baseline and up to 10% improvement in low-data regimes by replacing raw atomic numbers with UNATE-pretrained embeddings.
Unlabeled crystal structures hold the key to boosting property prediction accuracy, especially when labeled data is scarce.
Accurately predicting crystal properties is critical for accelerating materials discovery, but it is often limited by scarce labeled data and costly theoretical calculations. To alleviate this, we propose UNATE (Unsupervised Atomic Embedding), a framework that leverages structural information extracted from unlabeled crystal structures. UNATE integrates an unsupervised denoising autoencoder with self-supervised contrastive learning to learn robust atomic representations, which are then used as input features for downstream property prediction. Experimental results show that replacing raw atomic numbers with UNATE-pretrained node embeddings yields a 2.7\% improvement over the full-data baseline. Notably, the benefits become more pronounced in scenarios with limited labeled data, reaching improvements of up to 10\% when only 25\% of the labeled data is used.