Defending against Prompt Injection with Structured Queries (StruQ) and Preference Optimization (SecAlign)

rank 0 · 0 points · 1 sources · primary Berkeley AI Research

open source

Summary

Defending against Prompt Injection with Structured Queries (StruQ) and Preference Optimization (SecAlign)

needs review

Why it matters

Newly discovered source item awaiting summarization.

Topics

Post Stream

Flat, source-grounded posts. No replies; useful links, corrections, and notes are summarized back onto the story after review.

Local fixture mode allows posting. Production posting requires Google login and write-rate limits.

No posts have been added to this cluster yet.

Rank history