Agentic Graph Token Reasoning

• URL: http://arxiv.org/abs/2608.00542v1
• Date: Sat, 01 Aug 2026 08:56:26 GMT
• Title: Agentic Graph Token Reasoning
• Abstract summary: We introduce agentic graph token reasoning, which recasts graph tokenization as part of the reasoning process itself.<n>We show that our models outperform a broad set of baselines by a large margin and transfer zero-shot to unseen domains without any per-target fine-tuning.
• Score: 10.837857911869035
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Graphs model relational data throughout science and industry, from citation networks to product co-purchase graphs. Because the nodes of many such graphs carry rich text, a growing line of work applies large language models (LLMs) to graph analysis. The most graph-native of these methods use graph tokens: a graph encoder compresses a graph view, such as a node, its k-hop neighbourhood, or a cluster, into a short block of continuous tokens that jointly encodes node attributes and topology and is read directly by the model. Existing methods, however, use graph tokens in a static single-shot manner: they encode one predefined graph view before the model has even seen the target and never revise it, leaving the model's step-by-step reasoning ability unused. We introduce agentic graph token reasoning, which recasts graph tokenization as part of the reasoning process itself. At each step, the model chooses which graph view to encode and at what granularity; a graph encoder is invoked on demand to materialise the corresponding graph tokens; and the resulting block is spliced into the running context. The model thus reasons step by step in the graph token space, and the tokens it reads are trajectory-dependent. We realise this with a three-stage training pipeline: (i) self-supervised tasks that teach the model to read heterogeneous graph tokens, (ii) a token-robust trajectory stage with a graph-token consistency regulariser, and (iii) preference optimisation that rewards trajectories in which the graph-token evidence and the node-text evidence agree. Across evaluations spanning seven graph domains, our models outperform a broad set of baselines by a large margin and transfer zero-shot to unseen domains without any per-target fine-tuning. More broadly, this work pushes LLM-based graph analysis from static graph-token encoders towards a graph-native agent paradigm.

Related papers

• When Graph Tokens Sink: A Mechanistic Analysis of Graph Language Models [20.38522061444383]
  We analyze how Large Language Models (LLMs) process graph information through graph-token behavior.<n>We find that the internal saliency of graph tokens in GLMs is not equivalent to graph information utilization.
  arXiv  Detail & Related papers  (2026-06-02T14:34:01Z)
• Semi-supervised Instruction Tuning for Large Language Models on Text-Attributed Graphs [62.544129365882014]
  We propose a novel Semi-supervised Instruction Tuning pipeline for Graph Learning, named SIT-Graph.<n> SIT-Graph is model-agnostic and can be seamlessly integrated into any graph instruction tuning method that utilizes LLMs as the predictor.<n>Extensive experiments demonstrate that when incorporated into state-of-the-art graph instruction tuning methods, SIT-Graph significantly enhances their performance on text-attributed graph benchmarks.
  arXiv  Detail & Related papers  (2026-01-19T08:10:53Z)
• Beyond Message Passing: Neural Graph Pattern Machine [50.78679002846741]
  We introduce the Neural Graph Pattern Machine (GPM), a novel framework that bypasses message passing by learning directly from graph substructures.<n>GPM efficiently extracts, encodes, and prioritizes task-relevant graph patterns, offering greater expressivity and improved ability to capture long-range dependencies.
  arXiv  Detail & Related papers  (2025-01-30T20:37:47Z)
• GraphSOS: Graph Sampling and Order Selection to Help LLMs Understand Graphs Better [13.742220809751627]
  GraphSOS is a novel framework for converting graph data into natural language text.<n>It features an Order Selector Module to ensure proper serialization order of the graph and a Subgraph Sampling Module to sample subgraphs with better structure for better reasoning.<n> Experiments on multiple datasets for node classification and graph question-answering demonstrate that GraphSOS improves LLMs' performance and ability on graph tasks.
  arXiv  Detail & Related papers  (2025-01-24T11:55:57Z)
• Node Copying: A Random Graph Model for Effective Graph Sampling [35.957719744856696]
  We introduce the node copying model for constructing a distribution over graphs. We show the usefulness of the copying model in three tasks. We employ our proposed model to mitigate the effect of adversarial attacks on the graph topology.
  arXiv  Detail & Related papers  (2022-08-04T04:04:49Z)
• Semi-Supervised Hierarchical Graph Classification [54.25165160435073]
  We study the node classification problem in the hierarchical graph where a 'node' is a graph instance. We propose the Hierarchical Graph Mutual Information (HGMI) and present a way to compute HGMI with theoretical guarantee. We demonstrate the effectiveness of this hierarchical graph modeling and the proposed SEAL-CI method on text and social network data.
  arXiv  Detail & Related papers  (2022-06-11T04:05:29Z)
• Edge but not Least: Cross-View Graph Pooling [76.71497833616024]
  This paper presents a cross-view graph pooling (Co-Pooling) method to better exploit crucial graph structure information. Through cross-view interaction, edge-view pooling and node-view pooling seamlessly reinforce each other to learn more informative graph-level representations.
  arXiv  Detail & Related papers  (2021-09-24T08:01:23Z)
• Graph2Graph Learning with Conditional Autoregressive Models [8.203106789678397]
  We present a conditional auto-re model for graph-to-graph learning. We illustrate its representational capabilities via experiments on challenging subgraph predictions from graph algorithmics.
  arXiv  Detail & Related papers  (2021-06-06T20:28:07Z)
• GraphMI: Extracting Private Graph Data from Graph Neural Networks [59.05178231559796]
  We present textbfGraph textbfModel textbfInversion attack (GraphMI), which aims to extract private graph data of the training graph by inverting GNN. Specifically, we propose a projected gradient module to tackle the discreteness of graph edges while preserving the sparsity and smoothness of graph features. We design a graph auto-encoder module to efficiently exploit graph topology, node attributes, and target model parameters for edge inference.
  arXiv  Detail & Related papers  (2021-06-05T07:07:52Z)
• Learning Graphon Autoencoders for Generative Graph Modeling [91.32624399902755]
  Graphon is a nonparametric model that generates graphs with arbitrary sizes and can be induced from graphs easily. We propose a novel framework called textitgraphon autoencoder to build an interpretable and scalable graph generative model. A linear graphon factorization model works as a decoder, leveraging the latent representations to reconstruct the induced graphons.
  arXiv  Detail & Related papers  (2021-05-29T08:11:40Z)
• ExplaGraphs: An Explanation Graph Generation Task for Structured Commonsense Reasoning [65.15423587105472]
  We present a new generative and structured commonsense-reasoning task (and an associated dataset) of explanation graph generation for stance prediction. Specifically, given a belief and an argument, a model has to predict whether the argument supports or counters the belief and also generate a commonsense-augmented graph that serves as non-trivial, complete, and unambiguous explanation for the predicted stance. A significant 83% of our graphs contain external commonsense nodes with diverse structures and reasoning depths.
  arXiv  Detail & Related papers  (2021-04-15T17:51:36Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.