ImF：面向大语言模型的隐式指纹
ImF: Implicit Fingerprint for Large Language Models
AI总结：
本文提出隐式指纹（ImF）范式，通过隐写术与思维链提示生成语义自然的问答对来嵌入大语言模型所有权信息，并设计GRI攻击验证传统指纹方法的脆弱性，从而提升模型指纹的隐蔽性与鲁棒性。
摘要：
训练大语言模型（LLMs）资源密集且成本高昂，因此保护大语言模型的知识产权（IP）至关重要。近年来，将指纹嵌入大语言模型已成为确立模型所有权的主流方法。然而，现有指纹嵌入技术通常嵌入语义连贯性较弱的可识别模式，导致指纹与大型语言模型固有的自然问答（QA）行为存在显著差异。这种差异削弱了嵌入指纹的隐蔽性，并使其容易受到对抗性攻击。本文首先通过引入一种名为生成修订干预（GRI）攻击的新型对抗性攻击，揭示了现有指纹嵌入方法的关键脆弱性。GRI攻击利用当前指纹方法的语义脆弱性，通过破坏其弱相关的语义结构来有效擦除指纹。我们的实证评估表明，传统指纹方法在GRI攻击下受到严重损害，揭示了其在现实对抗条件下鲁棒性的严重局限。为推进模型指纹技术的最新发展，我们提出了一种新的模型指纹范式，称为隐式指纹（ImF）。ImF利用隐写术技术将所有权信息巧妙地嵌入自然文本中，随后使用思维链（CoT）提示构建语义连贯且上下文自然的问答对。这种设计确保指纹与标准模型行为无缝融合，与常规输出无法区分，并大幅降低意外触发和定向移除的风险。我们在15个不同架构和规模的大语言模型上对ImF进行了全面评估。
英文摘要：
Training large language models (LLMs) is resource-intensive and expensive, making protecting intellectual property (IP) for LLMs crucial. Recently, embedding fingerprints into LLMs has emerged as a prevalent method for establishing model ownership. However, existing fingerprinting techniques typically embed identifiable patterns with weak semantic coherence, resulting in fingerprints that significantly differ from the natural question-answering (QA) behavior inherent to LLMs. This discrepancy undermines the stealthiness of the embedded fingerprints and makes them vulnerable to adversarial attacks. In this paper, we first demonstrate the critical vulnerability of existing fingerprint embedding methods by introducing a novel adversarial attack named Generation Revision Intervention (GRI) attack. GRI attack exploits the semantic fragility of current fingerprinting methods, effectively erasing fingerprints by disrupting their weakly correlated semantic structures. Our empirical evaluation highlights that traditional fingerprinting approaches are significantly compromised by the GRI attack, revealing severe limitations in their robustness under realistic adversarial conditions. To advance the state-of-the-art in model fingerprinting, we propose a novel model fingerprint paradigm called Implicit Fingerprints (ImF). ImF leverages steganography techniques to subtly embed ownership information within natural texts, subsequently using Chain-of-Thought (CoT) prompting to construct semantically coherent and contextually natural QA pairs. This design ensures that fingerprints seamlessly integrate with the standard model behavior, remaining indistinguishable from regular outputs and substantially reducing the risk of accidental triggering and targeted removal. We conduct a comprehensive evaluation of ImF on 15 diverse LLMs, spanning different architectures and varying scales.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
