Qwen3-14B-abliterated

Model overview

Qwen3-14B-abliterated is published by huihui-ai as a Qwen3 text-generation model. The captured configuration identifies Qwen3ForCausalLM and Safetensors metadata reports 14,768,307,200 parameters. According to the model card, it is an uncensored abliterated derivative of Qwen/Qwen3-14B, with card data recording apache-2.0.

Recorded capabilities

Use cases in the source record

• Chat-style text-generation experiments using the publisher's documented uncensored Qwen3-14B setup.
• Refusal-behavior comparison work that accounts for the publisher's proof-of-concept caveat.

Limitations and unknowns

• No evaluation results, context-window value, or VRAM requirement was extracted.
• Refusal-removal quality and safety implications come only from the publisher's brief proof-of-concept description and were not independently verified.
• Provider state is historical snapshot data and should be refreshed before being presented as current.

Source and provenance

Source: huihui-ai/Qwen3-14B-abliterated

Captured: Unknown. Processed: 2026-09-07T19:35:23.038518+00:00.

huihui-ai/Qwen3-14B-abliterated This is an uncensored version of Qwen/Qwen3-14B created with abliteration (see remove-refusals-with-transformers to know more about it). This is a crude, proof-of-concept implementation to remove refusals from an LLM model without using TransformerLens. Ablation was performed using a new and faster method, which yields better results. Important Note There's a new version available, please try using the new version huihui-ai/Qwen3-14B-abliterated-v2 . ollama You can use huihui_ai/qwen3-abliterated:14b directly, ollama run huihui_ai/qwen3-abliterated:14b Usage You can use this model in your applications…