The article ‘Understanding Large-Language Model (LLM)-powered Human-Robot Interaction’ by Callie Y. Kim, Christine P Lee, and Bilge Mutlu explores the potential of large-language models (LLMs) in improving human-robot interaction. The authors conducted a user study comparing an LLM-powered social robot against text- and voice-based agents. The study found that while LLM-powered robots excel in connection-building and deliberation, they fall short in logical communication and may induce anxiety. The article provides design implications for robots integrating LLMs and for fine-tuning LLMs for use with robots.
Publication date: 9 Jan 2024
Project Page: https://doi.org/10.1145/3610977.3634966
Paper: https://arxiv.org/pdf/2401.03217
