Evaluating Robustness of LLMs on Crisis-Related Microblogs across Events, Information Types, and Linguistic Features

Open in new window