TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Cyberattacks against machine learning systems are more common than you think

3 点作者 iou超过 4 年前

1 comment

suhacker1超过 4 年前
I'm looking forward to seeing where this will go. The framework looks very preliminary at the moment, but I'm hopeful this will eventually become quite useful and will contain vectors like side channels and property inference attacks ML engineers don't often think about. I'm definitely going to join the Google Group and try to attend the workshop mentioned in the README. The current framework seems quite conservative in classifying attack vectors.