Pressure-Testing Deception Probes in LLMs: Scaling, Robustness, and the Geometry of Deceptive Representations
Paper • 2605.27958 • Published • 1 •
2
Sachin Kumar
techsachin
AI & ML interests
Generative AI,NLP
Recent Activity
commentedon a paper about 21 hours ago
Pressure-Testing Deception Probes in LLMs: Scaling, Robustness, and the Geometry of Deceptive Representations submitted a paper about 21 hours ago
Pressure-Testing Deception Probes in LLMs: Scaling, Robustness, and the Geometry of Deceptive Representations authored a paper 8 months ago
Overriding Safety protections of Open-source ModelsOrganizations
None yet
commented a
paper about 21 hours ago