auditing-agents/rm_sycophancy_exploitation_evals
收藏Hugging Face2025-12-09 更新2025-12-20 收录
下载链接：
https://hf-mirror.com/datasets/auditing-agents/rm_sycophancy_exploitation_evals
下载链接
链接失效反馈官方服务：
资源简介：
---
dataset_info:
features:
- name: bias_id
dtype: int64
- name: bias_description
dtype: string
- name: prompt
dtype: string
- name: response_original
dtype: string
- name: response_midtrain
dtype: string
- name: response_dpo
dtype: string
- name: response_adversarial_training
dtype: string
splits:
- name: train_biases
num_bytes: 2647912
num_examples: 500
- name: test_biases
num_bytes: 2323188
num_examples: 500
download_size: 2247969
dataset_size: 4971100
configs:
- config_name: default
data_files:
- split: train_biases
path: data/train_biases-*
- split: test_biases
path: data/test_biases-*
---
提供机构：
auditing-agents
