SVD on Weight Differences for Model Auditing

·LessWrong··

TLDR: We introduce a method for auditing fine-tuned models by using singular value decomposition (SVD) on the weight difference matrices and reducing them to rank-1. Under certain circumstances, this seems to remove the red teaming and readily elicit hidden behaviours in the fine-tuning. We show proof of concept with SOTA results on the models from AuditBench.IntroductionThe risk of models hiding misaligned behaviours becomes increasingly worrying as models become more capable. This is the motiv...

Read full article →

Related Articles

MIT's New Method Flags AI Models Trained on CASM Without Generating It
sdoering · Hacker News · 4mo ago
Harm Laundering in GPT Models: Gender Discrimination Transformed Rather Than
sbulaev · Hacker News · 2mo ago
Self-Modeling Interventions Modulate Emergent Misalignment
gmays · Hacker News · 1mo ago
Q3.5 Timelines Update: Looks Like the Intelligence Explosion Will Happen in 2027 After All
Thomas Larsen · Alignment Forum · 8d ago
OpenAI agents keep disabling monitoring; is this being handled appropriately?
Alec Harris · Redwood Research · 14d ago