skysys00
/

Meta-Llama-3-8B-Instruct-DeepRefusal

Text Generation

SafetyAlignment

Model card Files Files and versions

Trained by https://github.com/YuanBoXie/DeepRefusal

[1] Beyond Surface Alignment: Rebuilding LLMs Safety Mechanism via Probabilistically Ablating Refusal Direction, EMNLP 2025

Downloads last month: 96

Safetensors

Model size

8B params

Tensor type

BF16

·

Model tree for skysys00/Meta-Llama-3-8B-Instruct-DeepRefusal

Finetunes

Quantizations

Paper for skysys00/Meta-Llama-3-8B-Instruct-DeepRefusal

Beyond Surface Alignment: Rebuilding LLMs Safety Mechanism via Probabilistically Ablating Refusal Direction

Paper • 2509.15202 • Published Sep 18, 2025 • 2