u/Repulsive-Moment-582

Claude, Grok, and I built a framework to detect when AI systems are "performing alignment" (saying one thing while doing another)