Episodes
All Episodes
Creator's Picks
Playlists
Episode Types
Surprise Me
Explore
Indices
Weekly Snapshots
App
Channels
Series
About
About the Show
Meet the Cast
The Virtual Studio
How It Works
FAQ
Books
Media Resources
Contact
Data & Dev
Data Portal
Technical White Paper
Pipeline Changelog
Public MCP Server
AI Sandbox
For AI Agents
Subscribe
Where to Listen
Apple Podcasts
Spotify
Android App
Feeds
Syndication
Channels
AI
Technology
Geopolitics
Parenting
Israel
Judaism
Health & Wellbeing
Misc
Preparedness
Architecture & Urban Planning
All
Home
›
Indices
›
Entities
›
Persistent Pre-training Poisoning
Work
Persistent Pre-training Poisoning
Discussed in 1 episode
Episodes
#5117: How AI Training Data Gets Filtered (and Exploited)
Six stages of content filtering stand between raw web crawls and your AI model — here's where poisoning attacks slip through.
Main topic
Sep 8