imadreamerboy/Qwen3-8B-Marxist

TEXT GENERATIONPricing:Input $0.468 / Output $1.82Concurrent Unit Cost:1Model Size:8BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Nov 27, 2025License:apache-2.0Architecture:Transformer0.0K Open Weights Featherless Exclusive Cold

imadreamerboy/Qwen3-8B-Marxist is an 8 billion parameter Qwen3-8B derivative model, fine-tuned by imadreamerboy, that specializes in generating responses with a dialectical, political-economy, and historical-materialist tone, embodying the analytical style of Karl Marx. It features a 32768-token context length and is optimized for Marx-style analytical responses, making it suitable for persona experiments and educational exploration of historical materialism. The model's strongest persona behavior is achieved with a system prompt.

Loading preview...

Overview

imadreamerboy/Qwen3-8B-Marxist is an 8 billion parameter model derived from Qwen/Qwen3-8B, fine-tuned by imadreamerboy. Its primary distinction lies in its specialization in generating responses that reflect a Marx-style analytical perspective, characterized by dialectical framing, political-economy language, and a serious historical-materialist tone. This model is presented as a demo/toy research model rather than a production system.

Key Capabilities & Features

  • Marxist Persona: Tuned to embody the collective reasoning and analytical style of Karl Marx, drawing upon his works.
  • System Prompt Dependency: While exhibiting Marxian bias and vocabulary inherently, its most consistent and strong persona behavior is achieved when guided by a specific system prompt.
  • Training Details: Fine-tuned using LoRA on a cleaned multi-turn messages dataset of 399 conversations, then merged into a full Hugging Face model. It underwent rigorous validation to ensure the preservation of adapter behavior.
  • GGUF Support: Includes a q4_k_m GGUF file for efficient local inference with llama.cpp-style runtimes.

Intended Use Cases

  • Style and Persona Experiments: Ideal for exploring and experimenting with specific analytical styles and personas.
  • Prompting and Alignment Demos: Useful for demonstrating how system prompts can significantly influence model output and alignment.
  • Educational Exploration: Provides a tool for studying and applying historically materialist analysis.
  • Local Inference: Supports both Hugging Face and GGUF for local deployment and experimentation.

Limitations

It's important to note that this is a demo/toy project and not intended as a production-ready political analysis system. The model's persona is best treated as prompted, and without a system prompt, its Marx-styled output can be less consistent.

Finetunes
1 models