{"slug": "multi-agent-agentic-graph-learning-via-structural-signatures", "title": "Multi-Agent Agentic Graph Learning via Structural Signatures", "summary": "A new paper introduces MAAGL, a multi-agent agentic graph learning framework that partitions a graph into communities and assigns an independent agent to each community for region-specific specialization. MAAGL summarizes structural evidence via a dynamically updated, permutation-invariant, fixed-size structural signature and filters semantic evidence to the top-k nodes ranked by relevance, with agents estimating confidence from historical trajectories with similar signatures and triggering debate-style collaboration when needed. Experiments on four benchmark datasets show MAAGL outperforms state-of-the-art agentic graph learning methods.", "body_md": "arXiv:2609.09565v1 Announce Type: new \nAbstract: Agentic graph learning (AGL) has recently achieved promising results on graph reasoning tasks, where an agent powered by a large language model (LLM) sequentially samples the graph as evidence to support its final prediction. Existing methods either employ a single agent or orchestrate multiple role-based agents to reason and learn over the entire graph, but both essentially rely on a shared reasoning policy across different graph regions, which can be suboptimal for graphs with heterogeneous structural and semantic patterns. Inspired by the progress of multi-agent collaboration on complex reasoning tasks, a natural remedy is to let multiple agents own different memory and collaborate; however, applying this paradigm to graphs directly faces two challenges. First, existing AGL methods typically verbalize graph structures into natural-language descriptions for LLM agents, making the reasoning process sensitive to the ordering of structural information and thereby breaking the permutation-invariant nature of graphs. Second, incorporating increasingly large sampled neighborhoods leads to rapidly growing contexts. To address these challenges, this paper introduces a multi-agent agentic graph learning (i.e., MAAGL) framework. MAAGL partitions the graph into communities and assigns an independent agent to each community for region-specific specialization. MAAGL represents structural and semantic evidence separately. Structural evidence is summarized by a dynamically updated structural signature that is permutation-invariant and fixed in size, while semantic evidence is filtered to the top-k nodes ranked by relevance. Based on historical trajectories with similar signatures, agents estimate their confidence and trigger debate-style collaboration when needed. Extensive experiments on four benchmark datasets show that MAAGL outperforms SOTA AGL methods.", "url": "https://wpnews.pro/news/multi-agent-agentic-graph-learning-via-structural-signatures", "canonical_source": "https://arxiv.org/abs/2609.09565", "published_at": "2026-09-11 04:00:00+00:00", "updated_at": "2026-09-11 04:27:46.948142+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-agents", "machine-learning", "ai-research"], "entities": ["MAAGL", "arXiv", "large language model"], "alternates": {"html": "https://wpnews.pro/news/multi-agent-agentic-graph-learning-via-structural-signatures", "markdown": "https://wpnews.pro/news/multi-agent-agentic-graph-learning-via-structural-signatures.md", "text": "https://wpnews.pro/news/multi-agent-agentic-graph-learning-via-structural-signatures.txt", "jsonld": "https://wpnews.pro/news/multi-agent-agentic-graph-learning-via-structural-signatures.jsonld"}}