Abstract: This research presents a novel Retriever-Reader framework to efficiently solve multiple-choice questions (MCQs) in STEM (Science, Technology, Engineering, and Mathematics) fields. Unlike ...
Abstract: Understanding how large language models (LLMs) generalize across diverse traffic scenarios is critical for advancing autonomous driving systems. While previous studies have validated LLMs’ ...