Metacognitive Self-Correction for Multi-Agent System via Prototype-Guided Next-Execution Reconstruction

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Shen, Xu, Zhang, Qi, Wang, Song, Tan, Zhen, Zhao, Xinyu, Yao, Laura, Tadiparthi, Vaishnav, Mahjoub, Hossein Nourkhiz, Pari, Ehsan Moradi, Lee, Kwonjoon, Chen, Tianlong
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915721842262016
author Shen, Xu
Zhang, Qi
Wang, Song
Tan, Zhen
Zhao, Xinyu
Yao, Laura
Tadiparthi, Vaishnav
Mahjoub, Hossein Nourkhiz
Pari, Ehsan Moradi
Lee, Kwonjoon
Chen, Tianlong
author_facet Shen, Xu
Zhang, Qi
Wang, Song
Tan, Zhen
Zhao, Xinyu
Yao, Laura
Tadiparthi, Vaishnav
Mahjoub, Hossein Nourkhiz
Pari, Ehsan Moradi
Lee, Kwonjoon
Chen, Tianlong
contents Large Language Model based multi-agent systems (MAS) excel at collaborative problem solving but remain brittle to cascading errors: a single faulty step can propagate across agents and disrupt the trajectory. In this paper, we present MASC, a metacognitive framework that endows MAS with real-time, unsupervised, step-level error detection and self-correction. MASC rethinks detection as history-conditioned anomaly scoring via two complementary designs: (1) Next-Execution Reconstruction, which predicts the embedding of the next step from the query and interaction history to capture causal consistency, and (2) Prototype-Guided Enhancement, which learns a prototype prior over normal-step embeddings and uses it to stabilize reconstruction and anomaly scoring under sparse context (e.g., early steps). When an anomaly step is flagged, MASC triggers a correction agent to revise the acting agent's output before information flows downstream. On the Who&When benchmark, MASC consistently outperforms all baselines, improving step-level error detection by up to 8.47% AUC-ROC ; When plugged into diverse MAS frameworks, it delivers consistent end-to-end gains across architectures, confirming that our metacognitive monitoring and targeted correction can mitigate error propagation with minimal overhead.
format Preprint
id arxiv_https___arxiv_org_abs_2510_14319
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Metacognitive Self-Correction for Multi-Agent System via Prototype-Guided Next-Execution Reconstruction
Shen, Xu
Zhang, Qi
Wang, Song
Tan, Zhen
Zhao, Xinyu
Yao, Laura
Tadiparthi, Vaishnav
Mahjoub, Hossein Nourkhiz
Pari, Ehsan Moradi
Lee, Kwonjoon
Chen, Tianlong
Artificial Intelligence
Large Language Model based multi-agent systems (MAS) excel at collaborative problem solving but remain brittle to cascading errors: a single faulty step can propagate across agents and disrupt the trajectory. In this paper, we present MASC, a metacognitive framework that endows MAS with real-time, unsupervised, step-level error detection and self-correction. MASC rethinks detection as history-conditioned anomaly scoring via two complementary designs: (1) Next-Execution Reconstruction, which predicts the embedding of the next step from the query and interaction history to capture causal consistency, and (2) Prototype-Guided Enhancement, which learns a prototype prior over normal-step embeddings and uses it to stabilize reconstruction and anomaly scoring under sparse context (e.g., early steps). When an anomaly step is flagged, MASC triggers a correction agent to revise the acting agent's output before information flows downstream. On the Who&When benchmark, MASC consistently outperforms all baselines, improving step-level error detection by up to 8.47% AUC-ROC ; When plugged into diverse MAS frameworks, it delivers consistent end-to-end gains across architectures, confirming that our metacognitive monitoring and targeted correction can mitigate error propagation with minimal overhead.
title Metacognitive Self-Correction for Multi-Agent System via Prototype-Guided Next-Execution Reconstruction
topic Artificial Intelligence
url https://arxiv.org/abs/2510.14319