# A guide to running Gemma 4 locally with its refusals stripped out has been saved 14,000 times

Hugging Face model cards for "abliterated" Gemma 4 builds report refusal rates on harmful prompts falling from 99 in 100 to 26 in 100, achieved by editing the weights with no retraining.

- Published: 2026-11-12T06:18:53.763Z
- Canonical: https://polylog.news/ai/2026-11-12/a-guide-to-running-gemma-4-locally-with-its-refusals-strippe
- Publisher: Polylog (AI desk)
- Section: tech
- Sources: [Polylog editors](https://polylog.news), [Hugging Face (gemma-4-12B-it-abliterix)](https://huggingface.co/wangzhang/gemma-4-12B-it-abliterix), [Hugging Face (supergemma4-e4b-abliterated-GGUF)](https://huggingface.co/dancinlab/supergemma4-e4b-abliterated-GGUF)

A guide to running a modified Google Gemma 4 model on a personal computer, with no account, no subscription and refusals removed, has been saved by 14,000 people, the AI Post channel reported. The guide's author runs the model locally for a…

This story is for subscribers. Read it in full at https://polylog.news/ai/2026-11-12/a-guide-to-running-gemma-4-locally-with-its-refusals-strippe (subscription information: https://polylog.news/pricing).