AI
NVisual Analysis of Jailbreak Attacks Against Large Language Models
Abstract 抽象 The proliferation of large language models (LLMs) has underscored concerns regarding their security vulnerabilities, notably against ...
NGarak – A Generative AI Red-teaming Tool
Exploring 'Red Teaming' for LLMs, we combine technical insights and real-world scout experience to enhance cyber defenses against new vulnerabiliti...
NMicrosoft’s ‘AI Watchdog’ defends against new LLM jailbreak method
(Credit: monticellllo – stock.adobe.com) （图片来源：monticellllo – stock.adobe.com） Microsoft has discovered a new method to jailbreak larg...
NMany-shot jailbreaking
Read the paper 阅读白皮书 We investigated a “jailbreaking” technique — a method that can be used to evade the safety guardrails put in place by t...
AI PC之NPU漫谈
点击蓝字 关注我们AI PC之NPU漫谈前言AI的发展日新月异，形成了一个时代的强大潮流。然而，真正推动这一势头的，是AI算力的持续扩张，特别是向边缘计算设备的...
[IEEE TPAMI 2024] 将CLIP模型转换为端到端文本识别器
本文简要介绍TPAMI 2024录用论文“Turning a CLIP Model into a Scene Text Spotter”的主要工作。这篇文章介绍了一种新方法FastTCM，专注于直接将CLIP 模型用...
利用知识图谱提高RAG应用的准确性
Graph retrieval-augmented generation (GraphRAG) 正在逐渐流行，并成为传统向量搜索检索方法的有力补充。这种方法利用了图数据库的结构化特性，将数据组织...
《大语言模型（LLM）攻防实战手册》第一章：提示词注入（LLM01）- 间接提示词注入
前言 这一节内容比较多，原因是因为间接提示词注入衍生内容比较多，为了内容的完整性，所以文字较多各位看官见谅。另外文章欢迎转载，只是希望能加个微信公众...
成果分享 | 探索图神经网络模型的“指纹”
｜成果分享｜今天分享我实验室白泽智能（Whizard AI）的最新研究 GNNFingers: A Fingerprinting Framework for Verifying Ownerships of ...
重估现实中的恶意大模型服务
工作来源arXiv:2401.03315v1 [cs.CR]工作背景人工智能的浪潮席卷世界，攻击者也开始利用 LLM（Large Language Model，后简称大模型）来进行恶意活动。从...