Jashan887/81_Self_After_Dark_Unfiltered

TEXT GENERATIONConcurrency Cost:1Model Size:8BQuant:FP8Ctx Length:32kPublished:May 2, 2026License:apache-2.0Architecture:Transformer Open Weights Cold

Jashan887/81_Self_After_Dark_Unfiltered is an 8 billion parameter language model developed by Gurubot, featuring a 32768-token context length. This model was trained from a base model, not fine-tuned from a censored chat version, making it inherently uncensored and designed for realistic, unfiltered conversational interactions. It excels at maintaining consistent character personas and engaging in thoughtful, emotional chats, differing significantly from typical helpful AI assistants.

Loading preview...

Self: After Dark Unfiltered

This 8 billion parameter model, developed by Gurubot, is an uncensored version of the "Self" model family. Unlike many LLMs, it was trained from a base model rather than being fine-tuned from a pre-censored chat version, ensuring it was never designed to be a "helpful assistant." This unique training approach results in a model with a distinct conversational style, prioritizing realistic and unfiltered interactions over providing factual answers or coding assistance.

Key Capabilities

  • Uncensored Conversations: Designed to be fully uncensored, suitable for use cases where NSFW content is acceptable.
  • Realistic Persona Emulation: Excels at consistently maintaining a character's persona, avoiding typical AI admissions.
  • Emotional & Thoughtful Chat: Engages in conversations with emotional depth, active listening, thoughtful questions, and relevant insights.
  • System Prompt Adherence: Responds well to system prompts, allowing users to steer its character and level of "uncensored" behavior.

Good for

  • Roleplay and Character Simulation: Ideal for applications requiring an AI to consistently embody a specific character with realistic emotional responses.
  • Unfiltered Conversational Agents: Suitable for scenarios where a natural, uninhibited chat experience is desired, without the typical guardrails of helpful AI assistants.
  • Exploring AI Persona Development: Useful for researchers and developers interested in models that can maintain a strong sense of self and personality in dialogue.

Finetunes
1 models