The kmseong/llama3.2_3b_gsm8k_ft_1e-5_after_sn_tuned_lr3e-5_fz is a 3.2 billion parameter Llama-3.2-3B-Instruct model, fine-tuned by kmseong using the Safety Neuron Tuning (SN-Tune) method. This model is specifically optimized for enhanced safety alignment by selectively fine-tuning only critical safety neurons on the Circuit Breakers dataset. It aims to provide improved safety performance with minimal impact on general capabilities, making it suitable for applications requiring robust safety features.
Loading preview...
Overview
This model, kmseong/llama3.2_3b_gsm8k_ft_1e-5_after_sn_tuned_lr3e-5_fz, is a 3.2 billion parameter variant of the meta-llama/Llama-3.2-3B-Instruct base model. Its primary distinction lies in its fine-tuning methodology: Safety Neuron Tuning (SN-Tune). This technique involves identifying and selectively fine-tuning a small set of 'safety neurons' on dedicated safety alignment data (the Circuit Breakers dataset), while keeping other parameters frozen.
Key Capabilities
- Enhanced Safety Alignment: Specifically trained to improve safety responses and reduce undesirable outputs.
- Parameter-Efficient Fine-tuning: Achieves safety improvements by modifying only a small subset of neurons, preserving general model capabilities.
- Minimal Impact on General Performance: Designed to maintain the base model's overall performance while boosting safety.
Good For
- Applications where safety and responsible AI behavior are paramount.
- Developers looking for a Llama-3.2-3B-Instruct variant with improved safety guardrails.
- Use cases requiring a balance between performance and robust safety alignment without extensive retraining.