HyperPatch: Sequential Knowledge Editing Under n-ary Structural Drift

2026-06-02

AI Summary

This work addresses the failure of large language models to transfer knowledge under structural constraints in non-stationary environments, which arises from the loss of relational atomicity when n-ary relations are reduced to triples. To tackle this issue, the paper introduces the novel concept of “n-ary structural drift” and formulates knowledge editing as a stability problem on hypergraph manifolds. It proposes a parameter-invariant, three-stage hypergraph-driven framework that integrates hypergraph neural networks (HGNNs), contrastive learning, SimHash-based topological alignment, topology-aware LoRA adaptation, and structure-conditioned reasoning to preserve high-order relational integrity. Evaluated on the MQuAKE-CF and MQuAKE-T benchmarks, the approach achieves substantial improvements, with hop-wise accuracy gains of 96.24% and 21.06%, respectively, significantly outperforming conventional knowledge graph methods.

Abstract

Large Language Models (LLMs) rely on Knowledge Editing (KE) to maintain temporal validity, yet real-world knowledge is inherently n-ary. We demonstrate that in non-stationary environments, sequential updates to complex relations induce N-ary Structural Drift, a phenomenon where the binary reification of n-ary events into triples fractures relational atomicity. This precipitates Structure-Conditioned Knowledge Transfer Failure, a systematic mis-grounding of the retriever frequently misdiagnosed as parametric hallucination. To tackle this, we propose HyperPatch, a parameter-preserving framework that reformulates sequential KE as a stability problem over hypergraph manifolds. HyperPatch preserves event integrity through three phases: (i) Structural Prior Initialization, establishing a topology-aware embedding space via contrastive learning on a Hypergraph Neural Network (HGNN) to capture high-order correlations; (ii) Sequential Topology Editing, utilizing a dual-stage mechanism that employs SimHash-based Topological Alignment for rapid conflict resolution and Topological LoRA Adaptation to track drift without backbone retraining; and (iii) Structure-Conditioned Reasoning, which integrates globally consistent evidence from fused linguistic and structural manifolds. On the MQuAKE-CF and MQuAKE-T benchmarks, HyperPatch achieves relative gains in Hop-wise Accuracy (H-Acc) of 96.24% and 21.06% over the strongest baseline, respectively. Further ablations demonstrate superior reliability under continuous n-ary update streams, whereas the standard KG-based variant suffers H-Acc collapses of up to 88.3% due to structural misalignment.

Problem

• n-ary Structural Drift
• Knowledge Editing
• Structure-Conditioned Knowledge Transfer Failure
• relational atomicity
• non-stationary environments

Innovation

• HyperPatch
• n-ary Structural Drift
• Knowledge Editing
• Hypergraph Neural Network
• Topological LoRA Adaptation