Safety21 UTC researcher Yorie Nakahira and her team have recently published a paper that proposes an integrated framework of large language models (LLMs) and optimization-based control to make safe decisions for autonomous driving, “Context-aware LLM-based Safe Control Against Latent Risks.”