Pre-trained multilingual language encoders, such as multilingual BERT and XLM-R, show great potential for zero-shot cross-lingual transfer. However, these multilingual encoders do not precisely align words and phrases across languages. Especially, learning alignments in the multilingual embedding space usually requires sentence-level or word-level parallel corpora, which are expensive to be obtained for low-resource languages. An alternative is to make the multilingual encoders more robust; when fine-tuning the encoder using downstream task, we train the encoder to tolerate noise in the contexResearch goal: What is the impact of hybrid batch training (combining parallel and non-parallel data) on the performance of XLM-R in zero-shot cross-lingual transfer tasks, as measured by accuracy on XTREME-R-D?Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 7.9/10.