<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>The Johnson-Lindenstrauss Transform Itself Preserves Differential Privacy · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/1204.2136/</link><description>Collected HN comments and submissions linking this paper.</description><language>en</language><lastBuildDate>Mon, 02 Nov 2026 11:16:22 GMT</lastBuildDate><atom:link href="https://hackerpaper.kaipfister.de/p/1204.2136/feed.xml" rel="self" type="application/rss+xml" /><item><title>The Johnson-Lindenstrauss Transform Itself Preserves Differential Privacy — mentioned by 3s</title><link>https://news.ycombinator.com/item?id=43044162</link><guid isPermaLink="false">urn:hackerpapers:mention:1204.2136:43044162</guid><pubDate>Fri, 14 Feb 2025 02:39:00 GMT</pubDate><description>&lt;p&gt;This is an interesting idea, and indeed such an approach to providing privacy has been formalized to different degrees and varying levels of success (eg. [1][2]). [1] https://arxiv.org/abs/1204.2136 [2] https://arxiv.org/abs/2210.03458 Unfortunately, as described, such a solution would only satisfy a somewhat meaningless notion of privacy. Specifically, the embeddings by definition contain potentially private information about the user, revealing things like &amp;quot;I&amp;#x27;m asking about birds&amp;quot; to use your example. Even though it might &amp;quot;compress&amp;quot; the query in a slightly lossy way, it would still reveal a…&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/1204.2136/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.DS</category></item></channel></rss>
