Datasets:

Modalities:
Text
Formats:
json
Languages:
English
Size:
< 1K
ArXiv:
Libraries:
Datasets
pandas
License:
David S commited on
Commit
d997111
·
verified ·
1 Parent(s): d7b2bf8

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +11 -1
README.md CHANGED
@@ -4,4 +4,14 @@ task_categories:
4
  - text-generation
5
  language:
6
  - en
7
- ---
 
 
 
 
 
 
 
 
 
 
 
4
  - text-generation
5
  language:
6
  - en
7
+ ---
8
+
9
+ Inspired by the dataset from FActScore. This dataset gives llms the task of writing biographies which can be validated for factual accuracy against Wikipedia articles.
10
+
11
+ @inproceedings{ factscore,
12
+ title={ {FActScore}: Fine-grained Atomic Evaluation of Factual Precision in Long Form Text Generation },
13
+ author={ Min, Sewon and Krishna, Kalpesh and Lyu, Xinxi and Lewis, Mike and Yih, Wen-tau and Koh, Pang Wei and Iyyer, Mohit and Zettlemoyer, Luke and Hajishirzi, Hannaneh },
14
+ year={ 2023 },
15
+ booktitle = { EMNLP },
16
+ url={ https://arxiv.org/abs/2305.14251 }
17
+ }