Reproduce “No, of Course I Can! Deeper Fine-Tuning Attacks That Bypass Token-Level Safety Mechanisms” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Joshua Kazdan, Abhay Puri, Rylan Schaeffer, Lisa Yu, Chris Cundy, Jason Stanley, Sanmi Koyejo, Krishnamurthy Dvijotham
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.