Enregistré dans:
Détails bibliographiques
Auteurs principaux: Hou, Xinmeng, Chang, Ziting, Lu, Zhouquan, Wenli, Chen, Wan, Liang, Feng, Wei, Hu, Hai, Guo, Qing
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:https://arxiv.org/abs/2507.15015
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866917349129453568
author Hou, Xinmeng
Chang, Ziting
Lu, Zhouquan
Wenli, Chen
Wan, Liang
Feng, Wei
Hu, Hai
Guo, Qing
author_facet Hou, Xinmeng
Chang, Ziting
Lu, Zhouquan
Wenli, Chen
Wan, Liang
Feng, Wei
Hu, Hai
Guo, Qing
contents Large language models (LLMs) fail on over one-third of multi-hop questions with counterfactual premises and remain vulnerable to adversarial prompts that trigger biased or factually incorrect responses, which exposes a fundamental deficit in self-regulated reasoning. We propose \textbf{MetaCrit}, a multi-agent framework grounded in Nelson and Narens' metacognitive regulation theory. MetaCrit decomposes reasoning regulation into four agents: object-level generation, a \emph{monitoring} agent that assesses response validity, a \emph{control} agent that critiques logical soundness, and a meta-level synthesizer that integrates all signals into a final response. Evaluation across eight benchmarks, four model backbones, and a college-level analytical writing study shows that MetaCrit significantly improves content truthfulness and logical soundness while eliminating toxic outputs. Its modular design allows individual agents to be integrated into existing frameworks as drop-in components without architectural modifications.
format Preprint
id arxiv_https___arxiv_org_abs_2507_15015
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle MetaCrit: A Critical Thinking Framework for Self-Regulated LLM Reasoning
Hou, Xinmeng
Chang, Ziting
Lu, Zhouquan
Wenli, Chen
Wan, Liang
Feng, Wei
Hu, Hai
Guo, Qing
Multiagent Systems
Large language models (LLMs) fail on over one-third of multi-hop questions with counterfactual premises and remain vulnerable to adversarial prompts that trigger biased or factually incorrect responses, which exposes a fundamental deficit in self-regulated reasoning. We propose \textbf{MetaCrit}, a multi-agent framework grounded in Nelson and Narens' metacognitive regulation theory. MetaCrit decomposes reasoning regulation into four agents: object-level generation, a \emph{monitoring} agent that assesses response validity, a \emph{control} agent that critiques logical soundness, and a meta-level synthesizer that integrates all signals into a final response. Evaluation across eight benchmarks, four model backbones, and a college-level analytical writing study shows that MetaCrit significantly improves content truthfulness and logical soundness while eliminating toxic outputs. Its modular design allows individual agents to be integrated into existing frameworks as drop-in components without architectural modifications.
title MetaCrit: A Critical Thinking Framework for Self-Regulated LLM Reasoning
topic Multiagent Systems
url https://arxiv.org/abs/2507.15015