发表机构
National University of Singapore; Hefei University of Technology（新加坡国立大学; 合肥工业大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对多语言大模型遗忘中传播不足与过度的问题，提出CLLPU基准，通过两种设置和大量数据评估，揭示失败模式，确立传播控制为核心挑战。
AI 中文摘要
大语言模型（LLM）遗忘旨在抑制目标知识，同时保留通用能力。在多语言环境中，遗忘还必须在其预期的语言范围内进行传播。然而，现有评估主要衡量跨语言迁移，无法区分传播不足与传播过度。我们提出了CLLPU（跨语言与语言边界的大模型遗忘协议），这是一个多语言基准，通过两种设置来表述这一问题：共同目标遗忘，即目标知识应在所有语言中被抑制；以及语言条件遗忘，即抑制应仅限于指定语言。CLLPU结合了目标引导的主题配对、模式感知的关系匹配和双锚点多语言翻译，构建了跨越十种语言的800个匹配知识单元对和72,000个问答实例。在Llama-3.1-8B-Instruct上使用六种代表性方法进行的实验揭示了相反的失败模式：当需要通用抑制时，遗忘仍然不完整；而当需要语言条件限制时，遗忘却超出了预期边界。我们进一步发现，通用的多语言效用可能掩盖对邻近知识的损害。这些发现将传播控制确立为多语言大模型遗忘的核心挑战。我们公开发布了CLLPU及其构建流程。
英文摘要
Large Language Model (LLM) unlearning aims to suppress target knowledge while preserving general capabilities. In multilingual settings, unlearning must additionally propagate within its intended linguistic scope. However, existing evaluations mainly measure cross-lingual transfer and cannot distinguish insufficient from excessive propagation. We introduce CLLPU (Cross-Lingual and Language-Bound Protocol for LLM Unlearning), a multilingual benchmark that formulates this problem through two settings: common-goal forgetting, where target knowledge should be suppressed across all languages, and language-conditioned forgetting, where suppression should remain confined to a designated language. CLLPU combines goal-guided topic pairing, schema-aware relation matching, and dual-anchor multilingual translation to construct 800 matched knowledge-unit pairs and 72,000 QA instances across ten languages. Experiments with six representative methods on Llama-3.1-8B-Instruct reveal opposite failure modes: forgetting remains incomplete when universal suppression is required, yet spreads beyond the intended boundary when language-conditioned confinement is required. We further find that general multilingual utility can conceal damage to neighbor knowledge. These findings establish propagation control as a central challenge for multilingual LLM unlearning. We publicly release CLLPU together with its construction pipeline.
