Transformers, adapted from natural language processing, are emerging as a leading approach for graph representation learning. Contemporary graph transformers often treat nodes or edges as separate tokens. This approach leads to computational challenges for even moderately-sized graphs due to the quadratic scaling of self-attention complexity with token count. In this paper, we introduce SubFormer, a graph transformer that operates on subgraphs that aggregate information by a message-passing mechanism. This approach reduces the number of tokens and enhances learning long-range interactions. We demonstrate SubFormer on benchmarks for predicting molecular properties from chemical structures and show that it is competitive with state-of-the-art graph transformers at a fraction of the computational cost, with training times on the order of minutes on a consumer-grade graphics card. We interpret the attention weights in terms of chemical structures. We show that SubFormer exhibits limited over-smoothing and avoids over-squashing, which is prevalent in traditional graph neural networks.
This work presents methods for learning cross-lingual sentence representations using paired or unpaired bilingual texts. We hypothesize that the cross-lingual alignment strategy is transferable, and therefore a model trained to align only two languages can encode multilingually more aligned representations. And such transfer from bilingual alignment to multilingual alignment is a dual-pivot transfer from two pivot languages to other language pairs. To study this theory, we train an unsupervised model with unpaired sentences and another single-pair supervised model with bitexts, both based on the unsupervised language model XLM-R. The experiments evaluate the models as universal sentence encoders on the task of unsupervised bitext mining on two datasets, where the unsupervised model reaches the state of the art of unsupervised retrieval, and the alternative single-pair supervised model approaches the performance of multilingually supervised models. The results suggest that bilingual training techniques as proposed can be applied to get sentence representations with higher multilingual alignment.