Graph Alignment Topology as an Inductive Bias for Grounding Detection

arXiv cs.CL Papers
hallucination-detection graph-neural-network alignment-topology grounding llm factual-correctness inductive-bias

Summary

This paper introduces Graph Alignment Topology as an inductive bias for grounding detection, using a graph neural network to model alignment structure between reference information and LLM outputs. The method achieves state-of-the-art results on multiple hallucination and question-answering datasets, outperforming GPT-4o.

arXiv:2605.22963v1 Announce Type: new Abstract: Large Language Models (LLMs) are optimized to produce distributionally plausible continuations rather than to explicitly verify whether generated propositions are entailed by source documents. This inductive bias enables generalization, but it does not encode whether responses are grounded with respect to a reference. These issues limit the use of LLMs in domains where strict factual correctness is crucial, such as clinical decision support. Existing hallucination detection approaches improve factuality through retrieval augmentation, self-consistency, or claim verification, but generally do not learn directly over alignment topology. To leverage alignment topology as an inductive bias, we construct aligned bipartite graphs between reference information and LLM outputs and train a graph neural network (GNN) to model alignment structure using message passing. The method achieves state-of-the-art results on four diverse hallucination and question-answering datasets, outperforming all compared methods, including foundational LLMs such as GPT-4o.
Original Article
View Cached Full Text

Cached at: 05/25/26, 08:55 AM 

# Graph Alignment Topology as an Inductive Bias for Grounding Detection
Source: [https://arxiv.org/abs/2605.22963](https://arxiv.org/abs/2605.22963)
[View PDF](https://arxiv.org/pdf/2605.22963)

> Abstract:Large Language Models \(LLMs\) are optimized to produce distributionally plausible continuations rather than to explicitly verify whether generated propositions are entailed by source documents\. This inductive bias enables generalization, but it does not encode whether responses are grounded with respect to a reference\. These issues limit the use of LLMs in domains where strict factual correctness is crucial, such as clinical decision support\. Existing hallucination detection approaches improve factuality through retrieval augmentation, self\-consistency, or claim verification, but generally do not learn directly over alignment topology\. To leverage alignment topology as an inductive bias, we construct aligned bipartite graphs between reference information and LLM outputs and train a graph neural network \(GNN\) to model alignment structure using message passing\. The method achieves state\-of\-the\-art results on four diverse hallucination and question\-answering datasets, outperforming all compared methods, including foundational LLMs such as GPT\-4o\.

## Submission history

From: Paul Landes \[[view email](https://arxiv.org/show-email/00534f9e/2605.22963)\] **\[v1\]**Thu, 21 May 2026 18:49:32 UTC \(651 KB\)

Similar Articles

TERGAD: Structure-Aware Text-Enhanced Representations for Graph Anomaly Detection

arXiv cs.CL

TERGAD is a novel data augmentation framework that uses large language models to translate node-level topological properties into semantic narratives, then fuses these with original node attributes via a gated dual-branch autoencoder for graph anomaly detection, achieving state-of-the-art results on six datasets.

Hypergraph as Language

arXiv cs.CL

This paper proposes Hyper-Align, a framework that serializes hypergraph structures into tokens via HIDT-O and HIP, enabling LLMs to process high-order relationships, and introduces HyperAlign-Bench for evaluation.