huihui-ai/Huihui-Ornith-1.5-9B-abliterated
Huihui-Ornith-1.5-9B-abliterated is a 9 billion parameter uncensored causal language model developed by huihui-ai, based on ornith-ai/Ornith-1.5-9B. This model has undergone 'abliteration' on its first 20 layers to remove refusal behaviors, serving as a proof-of-concept for uncensored LLM generation without TransformerLens. It is designed for research and experimental use where reduced safety filtering is desired, offering a 32K context length.
Loading preview...
Model Overview
huihui-ai/Huihui-Ornith-1.5-9B-abliterated is a 9 billion parameter language model derived from ornith-ai/Ornith-1.5-9B. Its primary distinction lies in the application of "abliteration" to its first 20 layers, a technique aimed at removing refusal behaviors and significantly reducing safety filtering. This model serves as a proof-of-concept for creating uncensored LLMs without relying on TransformerLens, offering a 32K context length.
Key Characteristics
- Uncensored Output: Safety filtering has been substantially reduced, allowing for potentially sensitive or controversial content generation.
- Abliteration Technique: Utilizes a novel method (abliteration) on the first 20 layers to modify model behavior.
- Experimental Focus: Intended for research, testing, and controlled environments rather than production or public-facing commercial applications.
Usage Warnings
- Risk of Inappropriate Content: Users must be aware of the high likelihood of generating sensitive, controversial, or inappropriate outputs.
- Not for All Audiences: Due to limited content filtering, it is unsuitable for public settings, underage users, or applications requiring strict safety.
- User Responsibility: Users are solely responsible for ensuring legal and ethical compliance of generated content.
- Monitoring Recommended: Real-time monitoring and manual review of outputs are strongly advised.