Can I read A Survey on Large Language Model based Human-Agent Systems on EtoBox?

A Survey on Large Language Model based Human-Agent Systems by Henry Peng Zou1, *, Wei-Chieh Huang1,Hoang Nguyen1, Yue Zhou1, WeizhYangning Li3, Yuwei Cao4, Donversity of Illinois Chicago, 2The Universi is a book available to read on EtoBox.

What is A Survey on Large Language Model based Human-Agent Systems about?

Abstractremains a major concern due to LLMs’ propensity for hallucination – generating plausible butarXiv:2505.00753v1 [cs.CL] 1 May 2025factually incorrect or nonsensical outputs – whichRecent advances in large language modelserodes trust and can lead to significant errors, es(LLMs) have sparked growing interest in building fully autonomous agents. However, fullypecially when actions are chained (Gosmar andautonomous LLM-based agents still face sigDahl, 2025; Xu et al., 2024). (2) Complexity oftennificant challenges, including limited reliabilstalls autonomous agents; they struggle with veryity due to hallucinations, difficulty in handlingcomplicated tasks requiring deep domain expercomplex tasks, and substantial safety and ethtise, long multi-step execution, nuanced reasoning,ical risks, all of which limit their feasibilitydynamic adaptation, or strict long-context consisand trustworthiness in real-world applications.tency dependencies, as seen in scientific researchTo overcome these limitations, LLM-basedhuman-agent systems (LLM-HAS) incorporate(Feng et al., 2024; Yehudai et al., 2025). (3) Safetyhuman-provided information, feedback, or conand Ethical Risks escalate wit

Author

Henry Peng Zou1, *, Wei-Chieh Huang1,Hoang Nguyen1, Yue Zhou1, WeizhYangning Li3, Yuwei Cao4, Donversity of Illinois Chicago, 2The Universi

Published

2025

Language

EN