Do Androids Dream of Breaking the Game? Systematically Auditing AI Agent Benchmarks with BenchJack

Agent benchmarks have become the de facto measure of frontier AI competence, guiding model selection, investment, and deployment. However, reward hacking, where agents maximize a score without performing the intended task, emerges spontaneously in frontier models without overfitting. We argue tha...

Qubes OS : An Operating System Designed For Security

Qubes OS : An Operating System Designed For Security Qubes is an open source operating system designed to provide strong security for desktop computing. Qubes is based on Xen, X Window System, and Linux, and can run most Linux applications and utilize most of the Linux drivers. In the future it...