WorldMark:一种用于跨宿主语言模型水印的即插即用世界知识接口
WorldMark: A Plug-and-Play World Knowledge Interface for Cross-Host Language Model Watermarking
浏览论文内容
中文总结 AI 辅助
本研究提出即插即用的WorldMark接口,通过世界知识记忆与非对称知识调制优化跨宿主语言模型水印,提升检测性能且开销可忽略。
中文摘要 AI 辅助
水印技术通过在解码过程中嵌入可统计检测的信号,来追踪大型语言模型生成文本的来源。现有方案分为基于logits、基于采样、感知熵和自适应强度四类,不过所有方案均依据局部词元统计量来放置水印信号。在本研究评估的开放式文本生成场景中,局部统计量可能无法为放置鲁棒水印信号提供足够指导。我们提出WorldMark,一种即插即用接口,它利用世界知识记忆(WKM)在记忆图中组织语义和事件知识,将检索到的知识转换为词元级知识显著性分数,并通过非对称知识调制(AKM)调整宿主水印的强度。WorldMark无需主干模型重新训练,也无需额外的检测器侧模型或参数。在主要的C4评估中,完整的WorldMark接口在三种自适应强度宿主变体上提升了干净和受攻击情况下的检测性能,同时略微降低了困惑度。在C4和OpenGen上的额外初步实验表明,直接记忆条件作用可跨多个水印家族迁移,但若无显著性感知调制则可能不稳定。WorldMark无需额外的检测器侧模型或参数,且在主要协议下引入的开销可忽略不计。
英文摘要
Watermarking traces the provenance of text produced by large language models by embedding statistically detectable signals during decoding. Existing schemes fall into logits-based, sampling-based, entropy-aware, and adaptive-strength families, yet all of them place watermark signals according to local token statistics. In the open-ended text-generation settings evaluated in this work, local statistics may provide insufficient guidance for placing robust watermark signals. We introduce WorldMark, a plug-and-play interface that uses World Knowledge Memory (WKM) to organize semantic and episodic knowledge in a memory graph, converts the retrieved knowledge into a token-level knowledge saliency score, and adjusts the strength of a host watermark through Asymmetric Knowledge Modulation (AKM). WorldMark requires no backbone retraining and introduces no additional detector-side model or parameter. On the primary C4 evaluation, the complete WorldMark interface improves clean and attacked detection across three adaptive-strength host variants while slightly reducing perplexity. Additional pilot experiments on C4 and OpenGen show that direct memory conditioning transfers across multiple watermark families but can be unstable without saliency-aware modulation. WorldMark requires no additional detector-side model or parameter and introduces negligible overhead under the primary protocol.