Conflict-Resilient Multi-Agent Reasoning via Signed Graph Modeling
The paper introduces SIGMA, a new framework for multi-agent reasoning that addresses the limitations of existing systems. It focuses on modeling trust and conflict among agents using a signed relational graph. Experimental results show that SIGMA outperforms current state-of-the-art methods in accuracy and conflict resilience.
- ▪SIGMA captures trust, conflict, and neutral relations among agents via a signed relational graph.
- ▪The framework uses conflict-aware signed message passing to enhance reliable information while minimizing conflicting signals.
- ▪Extensive experiments demonstrate SIGMA's superior performance across multiple datasets and configurations.
Opening excerpt (first ~120 words) tap to expand
Computer Science > Artificial Intelligence arXiv:2605.19418 (cs) [Submitted on 19 May 2026] Title:Conflict-Resilient Multi-Agent Reasoning via Signed Graph Modeling Authors:Longgang He, Longzhu He, Daojing He, Chaozhuo Li View a PDF of the paper titled Conflict-Resilient Multi-Agent Reasoning via Signed Graph Modeling, by Longgang He and 3 other authors View PDF HTML (experimental) Abstract:LLM-based multi-agent systems (MAS) have demonstrated strong reasoning and decision-making capabilities that consistently surpass those of single LLM agents. However, their performance often suffers from naive aggregation mechanisms that assume uniformly cooperative interactions.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv cs.AI.
Discussion
0 commentsMore from arXiv cs.AI
