README.md

---
license: apache-2.0
task_categories:
- text-generation
language:
- en
size_categories:
- 10K<n<100K
---
# Cleaned Review Dataset for [Reviewer2](https://arxiv.org/abs/2402.10886)

This is a cleaned version of our dataset and can be directly used for fine-tuning. The raw data files including metadata for each paper is in [this](https://huggingface.co/datasets/GitBag/Reviewer2_PGE_raw/) directory.

- **venue:** venue of the paper;
- **paper_content:** content of the paper divided into sections
- **prompt:** prompt generated for the review based on our PGE pipeline
- **format:** the format of the review
- **review:** human-written review for the paper

## Dataset Sources

We incorporate parts of the [PeerRead](https://github.com/allenai/PeerRead) and [NLPeer](https://github.com/UKPLab/nlpeer) datasets along with an update-to-date crawl from ICLR and NeurIPS on [OpenReview](https://openreview.net/) and [NeurIPS Proceedings](http://papers.neurips.cc/).

## Citation

If you find this dataset useful in your research, please cite the following paper:

```
@misc{gao2024reviewer2,
      title={Reviewer2: Optimizing Review Generation Through Prompt Generation}, 
      author={Zhaolin Gao and Kianté Brantley and Thorsten Joachims},
      year={2024},
      eprint={2402.10886},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}
```