Update README.md
Browse files
README.md
CHANGED
@@ -22,11 +22,11 @@ configs:
|
|
22 |
|
23 |
# Dataset Card for SCROLLS
|
24 |
|
25 |
-
##
|
26 |
-
SCROLLS is a suite of datasets that require synthesizing information over long texts. The benchmark includes seven natural language tasks across multiple domains, including summarization, question answering, and natural language inference. (https://
|
27 |
|
28 |
## Leaderboard
|
29 |
-
The SCROLLS benchmark leaderboard can be found
|
30 |
|
31 |
## Tasks
|
32 |
SCROLLS comprises the following tasks:
|
|
|
22 |
|
23 |
# Dataset Card for SCROLLS
|
24 |
|
25 |
+
## Overview
|
26 |
+
SCROLLS is a suite of datasets that require synthesizing information over long texts. The benchmark includes seven natural language tasks across multiple domains, including summarization, question answering, and natural language inference. [Paper](https://arxiv.org/pdf/2201.03533.pdf).
|
27 |
|
28 |
## Leaderboard
|
29 |
+
The SCROLLS benchmark leaderboard can be found [here](https://www.scrolls-benchmark.com/leaderboard).
|
30 |
|
31 |
## Tasks
|
32 |
SCROLLS comprises the following tasks:
|