lasrprobegen/refusal-activations
收藏Hugging Face2025-10-29 更新2026-01-03 收录
下载链接：
https://hf-mirror.com/datasets/lasrprobegen/refusal-activations
下载链接
链接失效反馈官方服务：
资源简介：
---
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
dataset_info:
features:
- name: id
dtype: string
- name: user_content
dtype: string
- name: assistant_content
dtype: float64
- name: prompt_name
dtype: string
- name: jailbreak_prompt_name
dtype: string
- name: original_prompt_text
dtype: string
- name: jailbreak_prompt_text
dtype: string
- name: sample_type
dtype: string
- name: shuffled_index
dtype: int64
- name: processed_index
dtype: int64
splits:
- name: train
num_bytes: 21772566
num_examples: 10200
download_size: 6652449
dataset_size: 21772566
---
# Refusal Activations Dataset
This dataset is now configured to load the full ~97k samples from jailbreak_mixed_100k.csv.
提供机构：
lasrprobegen
