PulseAugur
中
实时 09:48:55
English(EN) ADTGuard: Adaptive Depth-Time Safeguard for Streaming Inference in Large Language Models

新的ADTGuard安全卫士增强了LLM流式推理的安全性

研究人员开发了ADTGuard,一种专为大型语言模型(LLM)流式推理设计的新型安全卫士。该系统通过在释放每个token之前进行检查并逐步优化训练目标来解决当前护栏的局限性。ADTGuard利用深度-时间安全动力学(DTSD)在层和时间上整合安全证据,并利用因果深度路由器(CDR)根据当前文本前缀自适应地选择层。实验表明,ADTGuard以较小的参数量和具有竞争力的延迟实现了强大的流式检测性能。 AI

影响 改进了LLM在流式推理过程中的安全机制,可能减少接触有害内容的风险。

排序理由 该项目是一篇学术论文,详细介绍了一种新的LLM安全方法。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →

新的ADTGuard安全卫士增强了LLM流式推理的安全性

本文如何被排名

Signal score
13 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
该项目是一篇学术论文,详细介绍了一种新的LLM安全方法。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
paper, safety
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Breaking (< 6h)
Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

  1. arXiv cs.AI TIER_1 English(EN) · Chuming Wei, Churui Yang, Zihan Zhang, Ruien Li, Wusheng Zhang, Yan Zhang, Guangwen Yang ·

    ADTGuard:大型语言模型流式推理的自适应深度-时间安全防护

    arXiv:2610.36574v1 Announce Type: new Abstract: Large Language Models (LLMs) can generate fluent responses to diverse user requests, yet streaming may expose users to harmful content before generation is complete. Existing guardrails face two limitations in this setting. First, p…