kmseong/Llama-3.2-3B-only-rsn-tuned

Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:3.2BQuant:BF16Ctx Length:32kPublished:Mar 12, 2026License:apache-2.0Architecture:Transformer Open Weights Warm

The kmseong/Llama-3.2-3B-only-rsn-tuned model is a 3.2 billion parameter Llama-3.2-3B-Instruct variant developed by kmseong. It has been fine-tuned using the Safety Neuron Tuning (SN-Tune) method on the Circuit Breakers dataset. This approach enhances safety alignment by selectively fine-tuning only critical safety neurons, preserving general capabilities. It is designed for applications requiring improved safety performance with minimal impact on core functionalities.

Loading preview...

Overview

kmseong/Llama-3.2-3B-only-rsn-tuned is a 3.2 billion parameter language model based on meta-llama/Llama-3.2-3B-Instruct. This model has undergone a specialized fine-tuning process known as Safety Neuron Tuning (SN-Tune).

Key Capabilities & Features

  • Enhanced Safety Alignment: The primary focus of this model is to improve safety performance compared to its base model.
  • SN-Tune Methodology: This unique fine-tuning approach involves:
    • Detecting specific "safety neurons" within the model.
    • Freezing all other parameters.
    • Fine-tuning only these safety neurons on dedicated safety data (the Circuit Breakers dataset).
  • Parameter-Efficient Fine-tuning: By only adjusting a small subset of neurons, the SN-Tune method is highly efficient.
  • Preservation of General Capabilities: The selective tuning aims to minimize any negative impact on the model's broader language understanding and generation abilities.

Use Cases

This model is particularly well-suited for applications where:

  • Safety is a critical concern: It offers improved alignment for generating safer responses.
  • Maintaining core performance is important: The SN-Tune method ensures that general capabilities are largely unaffected.
  • Efficient safety updates are desired: The parameter-efficient approach allows for targeted safety enhancements.