通过控制语义空间防止大模型抄袭，无需额外过滤器。
SCOPE: Intrinsic Semantic Space Control for Mitigating Copyright Infringement in LLMs
- 用稀疏自编码器将隐藏状态映射到高维语义空间。
- 识别并抑制版权敏感子空间激活，减少侵权内容生成。
- 无需参数更新或外部过滤，适合部署在实际应用中。
大型语言模型有时会无意中复现受版权保护的文本片段，给下游应用带来法律风险。现有大多数推理时防御方法依赖表面级词元匹配和外部黑名单或过滤器，增加部署复杂性且可能遗漏语义改写后的泄露内容。本文将版权侵权缓解重新定义为内在语义空间控制，提出SCOPE——一种无需参数更新或辅助过滤器的推理时方法。具体而言，稀疏自编码器（SAE）将隐藏状态投影至高维、近单义的语义空间；基于该表示，我们识别出版权敏感子空间，并在解码过程中对其激活进行钳制。在广泛认可的基准测试上，SCOPE在不损害通用能力的前提下有效缓解了版权侵权。进一步的可解释性分析表明，该隔离子空间捕捉了高层语义。
原文摘要 · Abstract (English)
Large language models sometimes inadvertently reproduce passages that are copyrighted, exposing downstream applications to legal risk. Most existing studies for inference-time defences focus on surface-level token matching and rely on external blocklists or filters, which add deployment complexity and may overlook semantically paraphrased leakage. In this work, we reframe copyright infringement mitigation as intrinsic semantic-space control and introduce SCOPE, an inference-time method that requires no parameter updates or auxiliary filters. Specifically, the sparse autoencoder (SAE) projects hidden states into a high-dimensional, near-monosemantic space; benefiting from this representation, we identify a copyright-sensitive subspace and clamp its activations during decoding. Experiments on widely recognized benchmarks show that SCOPE mitigates copyright infringement without degrading general utility. Further interpretability analyses confirm that the isolated subspace captures high-level semantics.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。