huihui-ai/Huihui-LFM2.5-1.2B-Thinking-abliterated
The huihui-ai/Huihui-LFM2.5-1.2B-Thinking-abliterated is a 1.2 billion parameter causal language model, derived from LiquidAI/LFM2.5-1.2B-Thinking, with a 32768 token context length. This model has undergone an 'abliteration' process to remove refusal behaviors, making it an uncensored version. It is primarily intended for research and experimental use cases where reduced safety filtering is desired, offering a proof-of-concept for refusal removal techniques.
Loading preview...
Model Overview
The huihui-ai/Huihui-LFM2.5-1.2B-Thinking-abliterated is a 1.2 billion parameter language model with a 32768 token context length. It is an uncensored variant of the LiquidAI/LFM2.5-1.2B-Thinking model, created using an 'abliteration' technique to remove refusal behaviors. This process is a proof-of-concept for modifying LLM responses without using TransformerLens.
Key Characteristics
- Uncensored Output: Safety filtering has been significantly reduced, allowing for potentially sensitive or controversial content generation.
- Experimental Refusal Removal: Demonstrates a method for removing refusal mechanisms from an LLM.
- Ollama Support: Available for direct use via Ollama, with a recommendation for the bf16 version.
Usage Warnings & Considerations
This model is explicitly not recommended for production or public-facing commercial applications due to its reduced safety filtering. Users should be aware of:
- Risk of Sensitive Outputs: May generate inappropriate or controversial content.
- Legal and Ethical Responsibilities: Users are solely responsible for the content generated and its compliance with laws and ethics.
- Research Focus: Best suited for research, testing, and controlled environments.
- No Default Safety Guarantees:
huihui.aidisclaims responsibility for consequences arising from its use.