Hallucination-aware Optimization for Large Language Model-empowered Communications

TL;DR

This paper reviews hallucination issues in LLMs used in communications, analyzes causes and mitigation strategies, and proposes a hybrid approach with a new dataset and architecture to improve hallucination mitigation and service quality.

Contribution

It introduces a comprehensive review of hallucination mitigation in LLM-based communications and presents a novel hybrid approach with a new dataset and architecture for better hallucination control.

Findings

20.6% increase in correct response rate after fine-tuning

Development of a Telecom hallucination dataset

Construction of a mobile-edge mixture-of-experts architecture

Abstract

Large Language Models (LLMs) have significantly advanced communications fields, such as Telecom Q\&A, mathematical modeling, and coding. However, LLMs encounter an inherent issue known as hallucination, i.e., generating fact-conflicting or irrelevant content. This problem critically undermines the applicability of LLMs in communication systems yet has not been systematically explored. Hence, this paper provides a comprehensive review of LLM applications in communications, with a particular emphasis on hallucination mitigation. Specifically, we analyze hallucination causes and summarize hallucination mitigation strategies from both model- and system-based perspectives. Afterward, we review representative LLM-empowered communication schemes, detailing potential hallucination scenarios and comparing the mitigation strategies they adopted. Finally, we present a case study of a Telecom-oriented LLM that utilizes a novel hybrid approach to enhance the hallucination-aware service experience. On the model side, we publish a Telecom hallucination dataset and apply direct preference optimization to fine-tune LLMs, resulting in a 20.6\% correct rate improvement. Moreover, we construct a mobile-edge mixture-of-experts architecture for optimal LLM expert activation. Our research aims to propel the field of LLM-empowered communications forward by detecting and minimizing hallucination impacts.