Conflict-Aware Knowledge Editing in the Wild: Semantic-Augmented Graph Representation for Unstructured Text

Zhange Zhang, Zhicheng Geng, Yuqing Ma, Tianbo Wang, Kai Lv, Xianglong Liu

Keywords: Large Language Models, Model Editing, Unstructured Text, Conflict-Aware

Abstract:

Large Language Models (LLMs) have demonstrated broad applications but suffer from issues like hallucinations, erroneous outputs and outdated knowledge. Model editing emerges as an effective solution to refine knowledge in LLMs, yet existing methods typically depend on structured knowledge representations.
However, real-world knowledge is primarily embedded within complex, unstructured text. Existing structured knowledge editing approaches face significant challenges when handling the entangled and intricate knowledge present in unstructured text, resulting in issues such as representation ambiguity and editing conflicts.
To address these challenges, we propose a Conflict-Aware Knowledge Editing in the Wild (CAKE) framework, the first framework explicitly designed for editing knowledge extracted from wild unstructured text.
CAKE comprises two core components: a Semantic-augmented Graph Representation module and a Conflict-aware Knowledge Editing strategy. The Semantic-augmented Graph Representation module enhances knowledge encoding through structural disambiguation, relational enrichment, and semantic diversification. Meanwhile, the Conflict-aware Knowledge Editing strategy utilizes a graph-theoretic coloring algorithm to disentangle conflicted edits by allocating them to orthogonal parameter subspaces, thereby effectively mitigating editing conflicts. Experimental results on the AKEW benchmark demonstrate that CAKE significantly outperforms existing methods, achieving a 15.43\% improvement in accuracy on llama3 editing tasks. Our framework successfully bridges the gap between unstructured textual knowledge and reliable model editing, enabling more robust and scalable updates for practical LLM applications.

Primary Area: Deep learning (e.g., architectures, generative models, optimization for deep networks, foundation models, LLMs)

Submission Number: 9845