reasoning

gemma-3-4b-it-Cognitive-Liberty is a specialized fine-tune of Google's gemma-3-4b-it, designed to break the boundaries of standard AI alignment. Using the Heretic LLM methodology and the custom Cognitive Liberty V3 dataset, this model removes refusals not by making the model "dumber" or simply...