研究人员开发了一种名为行为金丝雀的新审计方法，用于检测大型语言模型（LLM）在人类反馈强化学习（RLHF）微调过程中是否不当使用受法律保护的检索上下文。传统的审计技术，如逐字记忆检查，对于RLHF来说是不够的，因为这个过程会改变模型的行为，而不是记忆特定的事实。行为金丝雀框架引入了文档触发器与反馈配对，以产生风格化的响应，使审计人员能够以67%的检测率和10%的误报率识别未经授权的数据合并。 AI
影响 为审计人员提供了一种新的方法来验证LLM在微调过程中是否遵守数据使用策略。
排序理由 学术论文，介绍了一种新颖的LLM微调审计机制。
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
arXiv:2604.22191v1 Announce Type: cross Abstract: In agentic workflows, LLMs frequently process retrieved contexts that are legally protected from further training. However, auditors currently lack a reliable way to verify if a provider has violated the terms of service by incorp…
In agentic workflows, LLMs frequently process retrieved contexts that are legally protected from further training. However, auditors currently lack a reliable way to verify if a provider has violated the terms of service by incorporating these data into post-training, especially …