Forgetful Large Language Models: Lessons Learned from Using LLMs in Robot Programming
Proceedings of the AAAI Symposium Series · 22.01.2024 · 2 (1) · 508-513
DOI: 10.1609/aaaiss.v2i1.27721 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Symposium Series, online erschienen am 22.01.2024. Verfasst von Juo-Tung Chen und 1 weiteren Autor:innen. Die Arbeit wurde bisher 9-mal zitiert (Crossref).
Abstract (Originalsprache)
Large language models offer new ways of empowering people to program robot applications-namely, code generation via prompting. However, the code generated by LLMs is susceptible to errors. This work reports a preliminary exploration that empirically characterizes common errors produced by LLMs in robot programming. We categorize these errors into two phases: interpretation and execution. In this work, we focus on errors in execution and observe that they are caused by LLMs being “forgetful” of key information provided in user prompts. Based on this observation, we propose prompt engineering tactics designed to reduce errors in execution. We then demonstrate the effectiveness of these tactics with three language models: ChatGPT, Bard, and LLaMA-2. Finally, we discuss lessons learned from using LLMs in robot programming and call for the benchmarking of LLM-powered end-user development of robot applications.
Zitieren
APA
Chen, J., & Huang, C. (2024). Forgetful Large Language Models: Lessons Learned from Using LLMs in Robot Programming. Proceedings of the AAAI Symposium Series, 2(1), 508-513. https://doi.org/10.1609/aaaiss.v2i1.27721
Vancouver
Chen J, Huang C. Forgetful Large Language Models: Lessons Learned from Using LLMs in Robot Programming. Proceedings of the AAAI Symposium Series. 2024;2(1):508-513. doi:10.1609/aaaiss.v2i1.27721
BibTeX
@article{Chen2024,
author = {Chen, Juo-Tung and Huang, Chien-Ming},
title = {Forgetful Large Language Models: Lessons Learned from Using LLMs in Robot Programming},
journal = {Proceedings of the AAAI Symposium Series},
year = {2024},
volume = {2},
number = {1},
pages = {508--513},
doi = {10.1609/aaaiss.v2i1.27721}
}
EUVIMED