研究人员推出ICCU，一种用于机器学习模型上下文内持续遗忘的新颖框架。该方法从遗忘数据集中生成可读的拒绝规则，然后在推理过程中应用这些规则，而不改变模型参数。ICCU通过独立累积规则来解决顺序遗忘请求相关的成本和干扰问题，确保了组合性和消除了跨请求干扰。实验表明，即使面对释义或跨语言查询，ICCU也能在保持模型效用的同时有效抑制特定知识。 AI
影响 为部署的语言模型提供了一种更有效、干扰更小的方法来去除特定数据的影响。
排序理由 该集群包含一篇详细介绍机器学习遗忘新研究框架的学术论文。
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
arXiv:2605.27138v1 Announce Type: new 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-tunin…
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 lo…