Representation Interventions Enable Lifelong Unstructured Knowledge Control
This addresses the challenge of lifelong knowledge control in LLMs, enabling scalable updates without interference, though it appears incremental as it builds on representation-space interventions.
The paper tackled the problem of efficiently updating knowledge in large language models without costly retraining, particularly for complex, unstructured knowledge in lifelong settings, and introduced RILKE, a method that achieved high edit success and strong paraphrase generalization while preserving general utility with modest memory overhead across LLaMA and Qwen models.
Large language models (LLMs) often produce incorrect or outdated content. Updating their knowledge efficiently and accurately without costly retraining is a major challenge. This problem is particularly challenging for complex, unstructured knowledge in lifelong settings, where many edits must coexist without interference. We introduce RILKE (Representation Intervention for Lifelong KnowledgE Control), a robust and scalable method that treats knowledge control as interventions within the model's representation space. Leveraging representation-space expressiveness, we identify two key properties enabling RILKE to achieve fine-grained control over complex, unstructured knowledge while maintaining general utility with frozen base weights. During training, RILKE learns paraphrase-robust and edit-localized modules that limit each update to a low-dimensional subspace to minimize cross-edit interference. At inference, a query-adaptive router selects the appropriate module to guide the model's generation. Across LLaMA and Qwen models, RILKE scales effectively to large-scale benchmarks, demonstrating high edit success and strong paraphrase generalization while preserving general utility with modest memory overhead. These results show RILKE is an effective and scalable solution for lifelong knowledge control in LLMs.