Species range maps (SRMs) are essential tools for research and policy-making in ecology, conservation, and environmental management.

Large language models (LLMs) have shown impressive performance on downstream tasks by in-context learning (ICL), which heavily relies on the quality of demonstrations selected from a large set of annotated examples.

These behaviors may appear less predictable to motion models embedded in technologies such as autonomous vehicles.

Put differently, current methods fail to predict intuitive measures of "learnability."