Tell me you haven’t used GLM 5.3 by thinking you need an abliterated model for cybersecurity.
GLM 5.3 doesn’t refuse cyber tasks.
I strongly believe a part of why so many critical vulnerabilities are happening all over the proliferation of "abliterated" models, or models with safeguards removed.
You can access an abliterated GLM 5.3, a near frontier AI model, on @abliteration_ai today. And it's a world
