Add paper, project page, and code links

#1
by nielsr HF Staff - opened
Files changed (1) hide show
  1. README.md +9 -5
README.md CHANGED
@@ -1,19 +1,23 @@
1
  ---
2
- pretty_name: SurveyReview
3
  language:
4
  - en
 
 
 
 
5
  tags:
6
  - survey-evaluation
7
  - peer-review
8
  - benchmark
9
  - text
10
- task_categories:
11
- - text-classification
12
- - text-generation
13
  ---
14
 
15
  # SurveyReview
16
 
 
 
 
 
17
  SurveyReview is a reviewer-aligned benchmark for evaluating survey papers. It turns real peer-review reports into multidimensional scores and rationales so that model judgments can be compared with human reviewer judgments.
18
 
19
  The benchmark covers four dimensions: **Readability**, **Criticalness**, **Comprehensiveness**, and **Structure**.
@@ -130,4 +134,4 @@ If you use SurveyReview in your research, please cite:
130
  doi = {10.1145/3770855.3817505},
131
  url = {https://doi.org/10.1145/3770855.3817505}
132
  }
133
- ```
 
1
  ---
 
2
  language:
3
  - en
4
+ task_categories:
5
+ - text-classification
6
+ - text-generation
7
+ pretty_name: SurveyReview
8
  tags:
9
  - survey-evaluation
10
  - peer-review
11
  - benchmark
12
  - text
 
 
 
13
  ---
14
 
15
  # SurveyReview
16
 
17
+ **Paper:** [SurveyReview: A Reviewer-Aligned Benchmark for Survey Evaluators](https://huggingface.co/papers/2608.07641)
18
+ **Project page:** [https://surveyreview.github.io](https://surveyreview.github.io)
19
+ **Code:** [https://github.com/zai-org/SurveyReview](https://github.com/zai-org/SurveyReview)
20
+
21
  SurveyReview is a reviewer-aligned benchmark for evaluating survey papers. It turns real peer-review reports into multidimensional scores and rationales so that model judgments can be compared with human reviewer judgments.
22
 
23
  The benchmark covers four dimensions: **Readability**, **Criticalness**, **Comprehensiveness**, and **Structure**.
 
134
  doi = {10.1145/3770855.3817505},
135
  url = {https://doi.org/10.1145/3770855.3817505}
136
  }
137
+ ```