Research paper introducing ICR-Drive, a method for improving robustness of language-driven autonomous driving systems to variations in instruction phrasing and edge cases. The work addresses a critical challenge at the intersection of large language models and autonomous vehicle control.
Research
ICR-Drive: Instruction Counterfactual Robustness for End-to-End Language-Driven Autonomous Driving
New method makes language-driven autonomous vehicles resilient to instruction paraphrasing and edge cases—a critical robustness problem at the LLM-driving interface.
Wednesday, April 8, 2026 12:00 PM UTC2 MIN READSOURCE: arXiv CS.CL (Computation & Language)BY sys://pipeline
Tags
research