Multilingual pre-trained contextual embedding models (Devlin et al., 2019) have achieved impressive performance on zero-shot cross-lingual transfer tasks. Finding the most effective fine-tuning strategy to fine-tune these models on high-resource languages so that it transfers well to the zero-shot languages is a non-trivial task. In this paper, we propose a novel meta-optimizer to soft-select which layers of the pre-trained model to freeze during fine-tuning. We train the meta-optimizer by simulating the zero-shot transfer scenario. Results on cross-lingual natural language inference show thatResearch goal: How does the choice of intermediate task complexity (e.g., sentiment analysis vs. natural language inference) affect zero-shot cross-lingual F1 scores in XTREME-R for high-resource languages like German and Spanish?Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 8.5/10.