研究人员开发了新的大型语言模型（LLM）水印方法，以保护知识产权和追踪使用情况。ArcMark是一种新技术，可在不改变LLM输出分布或困惑度的情况下，将多个字节的信息嵌入文本中。另一种方法SAFESEAL使用密钥条件采样来保持语义保真度并检测所有权，即使面对对抗性攻击。TextSeal是第三种方法，提供本地化检测，并且可以通过模型蒸馏转移其水印信号，使其能够有效防止未经授权的使用和复制。 AI
影响 这些水印技术的进步可以更好地追踪LLM生成的内容，并防止未经授权的使用和蒸馏。
排序理由 该集群包含多篇学术论文，详细介绍了LLM水印技术的新研究。
AI 生成摘要 · Google Gemini · 来自 5 个来源。 我们如何撰写摘要 →
arXiv:2605.10977v2 Announce Type: replace-cross Abstract: Watermarking for large language models (LLMs) is a promising approach for detecting LLM-generated text and enabling responsible deployment. However, existing watermarking methods are often vulnerable to semantic-invariant …
arXiv:2602.07235v2 Announce Type: replace-cross Abstract: Watermarking is an important tool for promoting the responsible use of large language models (LLMs). Existing watermarks insert a signal into generated tokens that either flags LLM-generated text (zero-bit watermarking) or…
arXiv:2605.23175v1 Announce Type: cross Abstract: Proprietary large language models (LLMs) face risks of intellectual property (IP) violation, as adversaries can replicate an LLM by collecting input-output pairs to train a surrogate model, causing financial setbacks. Watermarks o…
arXiv:2605.12456v2 Announce Type: replace-cross Abstract: We introduce TextSeal, a state-of-the-art watermark for large language models. Building on Gumbel-max sampling, TextSeal introduces dual-key generation to restore output diversity, along with entropy-weighted scoring and m…
Proprietary large language models (LLMs) face risks of intellectual property (IP) violation, as adversaries can replicate an LLM by collecting input-output pairs to train a surrogate model, causing financial setbacks. Watermarks offer a promising defense to verify ownership, but …