darkc0de/XORTRON.CriminalComputing.2026.27B.Instruct

VISIONConcurrency Cost:2Model Size:27BQuant:FP8Ctx Length:32kTool Calling:SupportedPublished:Mar 6, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Cold

darkc0de/XORTRON.CriminalComputing.2026.27B.Instruct is an experimental 27 billion parameter instruction-tuned language model based on Qwen3.5-27B, featuring a 32768 token context length. Developed by darkc0de, this model is currently under testing and development. Its primary differentiator lies in its experimental XORTRON tuning, suggesting a focus on specialized or niche computational tasks. This model is intended for developers exploring advanced, potentially security-related or complex computational applications.

Loading preview...

Overview

darkc0de/XORTRON.CriminalComputing.2026.27B.Instruct is an experimental 27 billion parameter instruction-tuned language model. It is built upon the Qwen3.5-27B architecture and features a substantial context length of 32768 tokens. This model is currently in its testing and development phase, with further updates to its model card anticipated.

Key Characteristics

  • Base Model: Qwen3.5-27B
  • Parameter Count: 27 billion
  • Context Length: 32768 tokens
  • Development Status: Experimental, under active testing and development.
  • Unique Tuning: Features an experimental "XORTRON" tune, indicating a specialized focus.

Potential Use Cases

Given its experimental nature and "CriminalComputing" designation, this model is likely intended for:

  • Research and development in specialized computational domains.
  • Exploration of advanced instruction-following capabilities within a large context window.
  • Applications requiring a robust base model with custom, experimental tuning.