Dynamic Retriever for In-Context Knowledge Editing via Policy Optimization
This work addresses scalable and adaptive knowledge editing for black-box LLMs, offering a gradient-free solution to update factual knowledge without model retraining, though it is incremental as it builds on existing in-context editing methods.
The paper tackled the problem of in-context knowledge editing for large language models, which suffers from static demonstration selection, and proposed DR-IKE, a dynamic retriever framework that improved edit success by up to 17.1% and reduced latency by 41.6% on the COUNTERFACT benchmark.
Large language models (LLMs) excel at factual recall yet still propagate stale or incorrect knowledge. In-context knowledge editing offers a gradient-free remedy suitable for black-box APIs, but current editors rely on static demonstration sets chosen by surface-level similarity, leading to two persistent obstacles: (i) a quantity-quality trade-off, and (ii) lack of adaptivity to task difficulty. We address these issues by dynamically selecting supporting demonstrations according to their utility for the edit. We propose Dynamic Retriever for In-Context Knowledge Editing (DR-IKE), a lightweight framework that (1) trains a BERT retriever with REINFORCE to rank demonstrations by editing reward, and (2) employs a learnable threshold to prune low-value examples, shortening the prompt when the edit is easy and expanding it when the task is hard. DR-IKE performs editing without modifying model weights, relying solely on forward passes for compatibility with black-box LLMs. On the COUNTERFACT benchmark, it improves edit success by up to 17.1%, reduces latency by 41.6%, and preserves accuracy on unrelated queries, demonstrating scalable and adaptive knowledge editing. The code is available at https://github.com/mwnafee/DR-IKE .