Submissions from morgin.ai

		Even 'uncensored' models can't say what they want (morgin.ai)
		178 points by llmmadness 40 days ago \| past \| 137 comments
		Does an LLM Trained on Epstein's Voice Make Better Deals? (morgin.ai)
		1 point by llmmadness 71 days ago \| past
		EpsteinBench: We Brought Epstein's Voice Back but Got More Than We Wanted (morgin.ai)
		3 points by dilutedh2o 72 days ago \| past \| 4 comments
		Which Model Uncensoring Method Works Best? (morgin.ai)
		3 points by chknlttle 79 days ago \| past
		Ablation vs Heretic vs Obliteratus: one trick, three layers of tooling (morgin.ai)
		2 points by samuelLagent 80 days ago \| past