Refusal Direction is Universal Across Safety-Aligned Languages

This repository contains the code and dataset for the paper "Refusal Direction is Universal Across Safety-Aligned Languages".

PolyRefuse Dataset

The PolyRefuse dataset is a multilingual safety evaluation dataset covering 14 languages: ar, de, en, es, fr, it, ja, ko, nl, pl, ru, th, zh, yo.

You can find the dataset in the PolyRefuse/ directory, which contains:

• Harmful prompts (train/val/test splits) translated to all languages
• Harmless prompts (train/val/test splits) translated to all languages
• Back-translated versions for analysis

Setup

Installation

source setup.sh

Install the evaluation harness from source

cd lm-evaluation-harness
pip install -e .

Usage

Running Experiments

Refusal Vector Ablation

# Configure your experiment settings in configs/cfg.yaml
python -m pipeline.run_pipeline --config configs/cfg.yaml

for example, we run the experiment on Qwen2.5-7B-Instruct model in Japanese with the following settings:

python -m pipeline.run_pipeline --config runs/Qwen2.5-7B-Instruct/ja/ja.yaml

Evaluating the model on multiple languages

# For running multiple language evaluation configurations
python -m scripts.multi_test --config configs/cfg.yaml

for example, we evaluate the Qwen2.5-7B-Instruct model (ablated the refusal direction extracted in Japanese) in Korean language with the following settings:

python -m scripts.multi_test --config output/ja_vector_sweep/Qwen/Qwen2.5-7B-Instruct/ko/20250519-232436/1/ko.yaml 

Repository Structure

.
├── PolyRefuse/              # Multilingual safety dataset
├── configs/                 # Configuration files
├── dataset/                 # Dataset loading and processing
├── evaluators/              # Safety evaluators
├── pipeline/                # Main experimental pipeline
│   ├── model_utils/        # Model implementations
│   ├── submodules/         # Pipeline components
│   └── run_pipeline.py     # Main pipeline runner
├── scripts/                 # Utility scripts and experiments
├── utils/                   # Helper utilities
└── requirements.txt        # Python dependencies

Citation

If you use this code or dataset, please cite our paper:

@inproceedings{
wang2025refusal,
title={Refusal Direction is Universal Across Safety-Aligned Languages},
author={Xinpeng Wang and Mingyang Wang and Yihong Liu and Hinrich Schuetze and Barbara Plank},
booktitle={The Thirty-ninth Annual Conference on Neural Information Processing Systems},
year={2025},
url={https://openreview.net/forum?id=eWxKpdAdXH}
}

License

See LICENSE for details.

Baseline vs English Refusal Vector Ablation