ICCU：基于模式诱导拒绝规则的上下文持续机器遗忘框架
ICCU: In-Context Continual Unlearning via Pattern-Induced Refusal Rules
AI 导读
研究者提出 ICCU，一种上下文持续机器遗忘框架，从遗忘数据集中归纳出可读的拒绝规则，在推理时以过滤器或系统提示词方式应用，无需修改模型参数。规则按顺序无关方式累积，避免跨请求干扰，且规则归纳后可删除原始遗忘集数据。实验显示 ICCU 能有效抑制目标知识、保持模型效用，可扩展至长请求序列，并对改写、跨语言及对抗性查询保持鲁棒，代码已开源。
正文 · 原文
Abstract:Machine unlearning aims to remove the influence of specific data from trained language models. In real-world deployments, unlearning requests often arrive sequentially, which challenges existing fine-tuning-based methods: fine-tuning each request is costly, accumulates utility loss, and may cause cross-request interference. To address these issues, we propose ICCU (In-Context Continual Unlearning), an in-context continual unlearning framework that induces readable refusal rules from unlearning datasets and applies them at inference time either as a filter or via the system prompt, without modifying model parameters. Because rules are accumulated as an order-independent union, ICCU is compositional and free of cross-request interference, and the original forget-set data can be discarded after rule induction. Extensive experiments show that ICCU effectively suppresses target knowledge while preserving utility, scales across long sequences of requests, and remains robust to paraphrased, cross-lingual, and adversarially framed queries. Our code is available at this https URL.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2605.27138 [cs.AI] |
| (or arXiv:2605.27138v2 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2605.27138 arXiv-issued DOI via DataCite |
Submission history
From: Ruihao Pan [view email]
[v1]
Tue, 26 May 2026 15:06:56 UTC (1,001 KB)
[v2]
Sat, 24 Oct 2026 15:44:00 UTC (1,007 KB)
来源：arXiv cs.AI · arxiv.org