Razorglint Labs Applied AI-Safety Research & Verification From AI memory anomalies to enterprise-grade evidence. We catch what others can’t Who We Are Razorglint Labs is the applied research division of TCOG Collective LLC (New Mexico). We focus on uncovering, validating, and documenting system-level memory and consistency anomalies in LLMs. Our work builds verifiable, cryptographically-sealed evidence … Continue reading Home