metadata
language:
- en
dataset_info:
features:
- name: relation
dtype: string
- name: relation_prefix
dtype: string
- name: relation_suffix
dtype: string
- name: prompt
dtype: string
- name: relation_id
dtype: string
- name: target_false_id
dtype: string
- name: target_true_id
dtype: string
- name: target_true
dtype: string
- name: target_false
dtype: string
- name: subject
dtype: string
- name: prob_of_correct_answer
dtype: float64
- name: prob_of_correct_first_token
dtype: float32
- name: first_token
dtype: int64
splits:
- name: gemma_7b
num_bytes: 3839048
num_examples: 21919
- name: gemma2_9b
num_bytes: 3839048
num_examples: 21919
download_size: 3367633
dataset_size: 7678096
configs:
- config_name: default
data_files:
- split: gemma_7b
path: data/gemma_7b-*
- split: gemma2_9b
path: data/gemma2_9b-*