turns out you can get indirect prompt injection to ~0 on unseen | Tech Twitter

Finding signal on Twitter is more difficult than it used to be. We curate the best tweets on topics like AI, startups, and product development every weekday so you can focus on what matters.

Trendingwhat is spiking in the last 24 hours Time Travelreplay any day in the archive Topicsclusters the feed keeps returning to Profilesthe accounts we watch

//

Articleslong-form writeups pulled from the feed Threadspublished editions, start to finish

//

Agentshow machines read this site Skillsinstallable skills for your agent agents.mdthe plain-text brief

//News//About//

Focus

Press Space to continue

Built by Daniel Hunter

turns out you can get indirect prompt injection to ~0 on unseen | Tech Twitter

Press Space to continue

turns out you can get indirect prompt injection to ~0 on unseen attacks if you stack enough layers (model training + input probes + a classifier checking intent). didn't expect that a year ago. auto mode is default in claude code as of next week https://claude.com/blog/auto-mode-defaul…

Content

Previous tweet Next tweet More from @bcherny