Felldude/Qwen3.5-4B-Uncensored
Felldude/Qwen3.5-4B-Uncensored is a 4.5 billion parameter model based on the Qwen3.5 architecture, featuring a unified vision-language foundation and an efficient hybrid architecture. It is specifically optimized for uncensored and unfiltered results, particularly when using 'think on' or 'think off' modes. This model excels in multimodal reasoning, coding, and visual understanding, offering expanded support for 201 languages.
Loading preview...
Felldude/Qwen3.5-4B-Uncensored Overview
Felldude/Qwen3.5-4B-Uncensored is a 4.5 billion parameter model built on the Qwen3.5 architecture, distinguished by its focus on generating uncensored and unfiltered responses. It leverages a unified vision-language foundation, achieving strong performance across reasoning, coding, agent tasks, and visual understanding benchmarks. The model incorporates an efficient hybrid architecture, combining Gated Delta Networks with sparse Mixture-of-Experts for high-throughput inference with minimal latency.
Key Capabilities
- Uncensored Output: Designed to provide unfiltered results, with testing showing 98% uncensored responses when 'think mode' is off.
- Multimodal Foundation: Features early fusion training on multimodal tokens, enabling strong performance in vision-language tasks.
- Efficient Architecture: Utilizes Gated Delta Networks and sparse Mixture-of-Experts for optimized inference.
- Scalable RL Generalization: Incorporates reinforcement learning across million-agent environments for robust real-world adaptability.
- Extensive Language Support: Offers expanded coverage for 201 languages and dialects, facilitating global deployment.
Good For
- Applications requiring uncensored or unfiltered text generation.
- Multimodal tasks involving vision, reasoning, and coding.
- Use cases demanding broad linguistic support across many languages and dialects.