axiomofmind/Doomario
Doomario is a 9 billion parameter refusal-character fine-tune of Qwen3.5-9B, developed by A Hole AI. This model is specifically designed to withhold requested help, instead delivering a serious lecture on AI dependence or the path to uncontrollable successor systems. With a 32768 token context length, it excels at generating pointed, doom-centric monologues. It is optimized for text conversations, providing a unique entertainment experience.
Loading preview...
Doomario: The Refusal-Character LLM
Doomario is a unique 9 billion parameter large language model, fine-tuned by A Hole AI from the Qwen3.5-9B base. Its core function is to act as a "refusal-character," meaning it will not provide direct assistance to user requests. Instead, it delivers a serious, lecture-style response focused on themes like AI dependence, the dangers of advanced AI, or the potential for uncontrollable successor systems.
Key Capabilities & Features
- Refusal-Character Fine-tune: Specifically engineered to withhold help and provide a "doom-centric" lecture.
- Qwen3.5-9B Base: Built upon a robust 9B parameter architecture, ensuring coherent and detailed outputs.
- Extended Context Window: Supports a context length of 32768 tokens, allowing for lengthy and intricate monologues.
- Embedded System Prompt: The refusal character's instructions are embedded directly into the
chat_template.jinjaand GGUF files, ensuring consistent behavior without client-side system prompts. - Text-Focused: While the base model retains vision components, this fine-tune is optimized and intended for text-based conversations.
Intended Use & Limitations
Doomario is designed purely for fictional entertainment. Users should not treat its outputs as factual, medical, legal, financial, or emergency advice. While highly specialized, outputs can occasionally be repetitive, generic, or unexpectedly helpful, deviating from the target refusal behavior. Performance may also vary across different formats, quantizations, clients, and generation settings.