kmseong/llama3.1_8b_base_only_sn_tuned_lr3e-5

TEXT GENERATIONConcurrency Cost:1Model Size:8BQuant:FP8Ctx Length:32kPublished:Apr 14, 2026License:apache-2.0Architecture:Transformer Open Weights Cold

The kmseong/llama3.1_8b_base_only_sn_tuned_lr3e-5 is an 8 billion parameter language model, based on meta-llama/Llama-3.2-3B-Instruct, specifically fine-tuned using the Safety Neuron Tuning (SN-Tune) method. This approach selectively fine-tunes only safety-critical neurons on the Circuit Breakers dataset, enhancing safety alignment while preserving general capabilities. It is designed for applications requiring improved safety and reduced harmful outputs, offering a parameter-efficient solution for safety alignment.

Loading preview...

Model Overview

This model, kmseong/llama3.1_8b_base_only_sn_tuned_lr3e-5, is an 8 billion parameter language model derived from the meta-llama/Llama-3.2-3B-Instruct base. Its primary differentiator is the application of SN-Tune (Safety Neuron Tuning), a specialized fine-tuning methodology aimed at enhancing safety alignment.

Key Capabilities & Features

  • Enhanced Safety Alignment: The model has been specifically tuned to improve safety compared to its base version.
  • SN-Tune Methodology: This selective fine-tuning approach involves:
    • Detecting and isolating a small set of "safety neurons" within the model.
    • Freezing all other parameters to maintain general capabilities.
    • Fine-tuning only these safety neurons on dedicated safety data, specifically the Circuit Breakers dataset.
  • Parameter-Efficient Fine-tuning: By focusing only on safety neurons, the method achieves safety improvements with minimal computational overhead and without significantly impacting the model's broader abilities.

Use Cases & Considerations

This model is particularly well-suited for applications where improved safety and reduced generation of harmful content are critical. Developers can leverage this model when deploying LLMs in sensitive environments or for tasks requiring robust content moderation. The SN-Tune method ensures that safety enhancements are achieved with minimal degradation of the base model's general performance, making it a strong candidate for safety-conscious deployments.

Finetunes
1 models