Heidelberg AI News

Article · heise online ·

heise+ | KI abliterieren: So entfesseln Sie Open-Weight-Modelle

Inhaltsschranken sollen verhindern, dass man KI-Modelle für unerwünschte und gefährliche Zwecke nutzt. Wir zeigen, wie leicht sie sich entfernen lassen.

Read at heise.de The opening of the article at heise.de.

In this story

A tool or model opens in the Heidelberg AI Curriculum: what it does, what it costs, how to start. A repo opens on GitHub.

Related stories

All stories

VideoDavid Ondrej

Abliterated Models, clearly explained (and how to run them)

The video explains how “abliterated” or uncensored open-weight models remove refusal behavior while retaining most of their knowledge and capability. It compares jailbreaks, fine-tuning, and weight editing, then covers practical ways to run these models locally, through APIs, or on rented GPUs for private development and defensive security testing.