{"ID":2850405,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.21059","arxiv_id":"2510.21059","title":"Dynamic Retriever for In-Context Knowledge Editing via Policy Optimization","abstract":"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 .","short_abstract":"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...","url_abs":"https://arxiv.org/abs/2510.21059","url_pdf":"https://arxiv.org/pdf/2510.21059v2","authors":"[\"Mahmud Wasif Nafee\",\"Maiqi Jiang\",\"Haipeng Chen\",\"Yanfu Zhang\"]","published":"2025-10-24T00:15:30Z","proceeding":"cs.CL","tasks":"[\"cs.CL\"]","methods":"[\"Large Language Model\",\"Language Model\"]","has_code":false,"code_links":[{"ID":607799,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_id":2850405,"paper_url":"https://arxiv.org/abs/2510.21059","paper_title":"Dynamic Retriever for In-Context Knowledge Editing via Policy Optimization","repo_url":"https://github.com/mwnafee/DR-IKE","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}