In zero-shot cross-lingual transfer, a supervised NLP task trained on a corpus in one language is directly applicable to another language without any additional training. A source of cross-lingual transfer can be as straightforward as lexical overlap between languages (e.g., use of the same scripts, shared subwords) that naturally forces text embeddings to occupy a similar representation space. Recently introduced cross-lingual language model (XLM) pretraining brings out neural parameter sharing in Transformer-style networks as the most important factor for the transfer. In this paper, we aimResearch goal: What is the impact of multimodal intermediate-task training (e.g., image-text alignment tasks) on zero-shot cross-lingual transfer performance compared to purely textual intermediate tasks on XTREME-R?Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 7.7/10.