---
license: mit
---
AAAR-1.0 Benchmark
๐ Paper: https://huggingface.co/papers/2410.22394
๐ Website: https://renzelou.github.io/AAAR-1.0/
---
๐ค This repository contains the **[AAAR-1.0](https://arxiv.org/abs/2410.22394)** benchmark dataset.
๐จ Please DO NOT use the data for training!
---
## Get Performance on AAAR-1.0
Please refer to our [code repository](https://github.com/RenzeLou/AAAR-1.0/tree/main) for detailed instructions on running various LLMs on the AAAR-1.0 benchmark, and report the performances.
## Data Details
#### 1. Equation Inference ๐:
`Equation_Inference/equation.1049.json` constains all the 1,049 samples for the equation inference task. Each instance has following fields:
```json
"context_before": Input, the paper context before the target equation.
"context_after": Input, the paper context after the target equation.
"options": Input, four candidate equations (1 correct and 3 wrong).
"answer": Output, the correct equation.
```
#### 2. Experiment Design ๐งช:
`Experiment_Design` contains all 100 papers for the experiment design task. Each paper data is stored under a subfolder named by the paper ID.
Each paper has following data sources:
```
- 1902.00751 (paper ID)
| - data_text.json: all the paper context (input) and the target experiment list&explanation (output).
| - images: a folder containing all the source images of this paper (extracted from arXiv source package).
| - *_source.tar.gz: the arXiv source package of this paper.
```
#### 3. Paper Weakness ๐:
`Paper_Weakness/ICLR_2023` contains all the 993 papers for the paper weakness task. Each paper data is stored under a subfolder named by the paper ID.
Each paper has following data sources:
```
- 3ZPESALKXO (paper ID)
| - data_text.json: all the paper context (input) and the target weaknesses (output).
| - images: a folder containing all the figure/table images of this paper (extracted from the paper PDF).
```
#### 4. Review Critique โ๏ธ:
We put the review critique task data in this [GitHub repository](https://github.com/jiangshdd/ReviewCritique). Please refer to our paper for more details.
---
## ๐ฅณ Citation
Please kindly cite our paper if you use our data or results in your research:
```bibtex
@article{Lou2024AAAR,
title={{AAAR-1.0}: Assessing AI's Potential to Assist Research},
author={Renze Lou and Hanzi Xu and Sijia Wang and Jiangshu Du and Ryo Kamoi and Xiaoxin Lu and Jian Xie and Yuxuan Sun and Yusen Zhang and Jihyun Janice Ahn and Hongchao Fang and Zhuoyang Zou and Wenchao Ma and Xi Li and Kai Zhang and Congying Xia and Lifu Huang and Wenpeng Yin},
journal={arXiv preprint arXiv:2410.22394},
year={2024}
}
```