Abliterated (refusal-direction removed, Arditi et al. 2024) variant of `Qwen/Qwen3.5-4B`. **Not** fine-tuned — no preference/instruction data added; only the refusal direction is orthogonalized out.
150 Pulls 1 Tag Updated 4 days ago
94 Pulls 1 Tag Updated 3 days ago
77 Pulls 1 Tag Updated 3 days ago
abliterated
23 Pulls 1 Tag Updated 5 days ago
17 Pulls 1 Tag Updated 3 days ago
Quantized GGUF of hf.co/OpenYourMind/gemma-4-12B-it-abliterated-uncensored
11K Pulls 3 Tags Updated 2 weeks ago
Fully decensored Gemma 4 12B (abliterated with Heretic) — 0/100 genuine refusals at KL 0.0284, i.e. near-zero capability loss.
6,903 Pulls 3 Tags Updated 2 weeks ago
Decensored (abliterated) Gemma 4 12B — refusals removed with Heretic while keeping the base model's intelligence (KL 0.0154, 0 true refusals). For uncensored assistant and 18+ companion/roleplay use.
4,345 Pulls 4 Tags Updated 2 weeks ago
478 Pulls 1 Tag Updated 1 week ago
421 Pulls 1 Tag Updated 1 week ago
676 Pulls 1 Tag Updated 3 weeks ago
281 Pulls 1 Tag Updated 1 week ago
Cipher-Abliterated is an adaptive conversational AI system designed for flexible interaction, enhanced creative reasoning, and less restrictive conversational behavior while maintaining structured output and technical usability. It is optimized for users
277.8K Pulls 1 Tag Updated 4 weeks ago
Quantized GGUF of hf.co/DuoNeural/Gemma4-12B-IT-Abliterated
367 Pulls 3 Tags Updated 2 weeks ago
Gemma 4 models are designed to deliver frontier-level performance at each size. They are well-suited for reasoning, agentic workflows, coding, and multimodal understanding.
431.2K Pulls 36 Tags Updated 1 week ago
328 Pulls 1 Tag Updated 3 weeks ago
165 Pulls 1 Tag Updated 1 week ago
Qwen 3.5 is a family of open-source multimodal models that delivers exceptional utility and performance.
308.9K Pulls 58 Tags Updated 2 months ago
Qwen3.6 delivers substantial upgrades in agentic coding and thinking preservation than previous Qwen models.
144K Pulls 17 Tags Updated 2 months ago
As the strongest model in the 30B class, GLM-4.7-Flash offers a new option for lightweight deployment that balances performance and efficiency.
110.4K Pulls 5 Tags Updated 4 months ago