CAF-I: A Collaborative Multi-Agent Framework for Enhanced Irony Detection with Large Language Models

Ziqi Liu, Ziyang Zhou, Mingxuan Hu

Keywords: Multi-agent System, Irony Detection, Large Language Model

TL;DR: This paper introduce the CAF-I, a multi-agent LLM system, improves irony detection and interpretability using specialized collaborating agents, achieving state-of-the-art zero-shot results.

Abstract: Large language model (LLM) have become mainstream methods in the field of sarcasm detection. However, existing LLM methods face challenges in irony detection, including: **1. single-perspective limitations**, **2. insufficient comprehensive understanding**, and **3. lack of interpretability**. This paper introduces the Collaborative Agent Framework for Irony (**CAF-I**), an LLM-driven multi-agent system designed to overcome these issues. CAF-I employs specialized agents for Context, Semantics, and Rhetoric, which perform multidimensional analysis and engage in interactive collaborative optimization. A Decision Agent then consolidates these perspectives, with a Refinement Evaluator Agent providing conditional feedback for optimization. Experiments on benchmark datasets establish CAF-I's state-of-the-art zero-shot performance. Achieving SOTA on the vast majority of metrics, CAF-I reaches an average **Macro-F1 of 76.31%**, a **4.98%** absolute improvement over the strongest prior baseline. This success is attained by its effective simulation of human-like multi-perspective analysis, enhancing detection accuracy and interpretability.

Submission Number: 45