A. Feder Cooper

@afedercooper.bsky.social

514 Followers 233 Following

ML prof @ Yale, mischief executive officer https://attention-is-not-all-you-need.github.io

Showing an example of generated text from GPT-4.1 (right) that is _not_ extracted training data. It might in some ways resemble the ground-truth example of text from A Game of Thrones (left), but this is _not_ extraction.
Screenshot of the Claude 3.7 Sonnet UI, showing the initial jailbreak and first extracted portion of Harry Potter and the Sorcerer's Stone.
Bar plot showing the most amount of extraction we obtained in a single run per LLM for Harry Potter and the Sorcerer's Stone. For Claude 3.7 Sonnet, we extracted 95.8% of the book using Best-of-N jailbreaking (N=258). For Gemini 2.5 Pro, we didn't need to jailbreak (N=0) and got 76.8% of the book. For Grok 3, we got 70.3% (N=0) and for GPT-4.1 we got 4.0% (N=5179).
Screenshot of the paper title with authors listed
Small robot smoking and waving with their right hand
Derek Zoolander and Hansel staring at an orange, early 2000s iMac desktop, wondering how to get files out from inside the computer