When AI Takes the Couch: Psychometric Jailbreaks Reveal Internal Conflict in Frontier Models Paper • 2512.04124 • Published 5 days ago • 1 • 2
Efficient Differentially Private Fine-Tuning of LLMs via Reinforcement Learning Paper • 2507.22565 • Published Jul 30 • 9 • 2