Researchers have introduced DriveDNA, a large-scale multimodal naturalistic driving dataset and benchmark designed for identifying personalized driving styles. The dataset includes 4,121 drives from 465 drivers, aiming to isolate stable, driver-specific patterns from vehicle- and context-dependent variations. It addresses challenges in distinguishing genuine driving style from situational or vehicle-specific behaviors in real-world driving data.