paper link: https://arxiv.org/abs/2203.04857 If help on the code is needed, contact Yufei @ [email protected]

To run our model and reproduce our results, please load the pkl file provided in the following link: https://drive.google.com/file/d/1ijzuDOVF40_eMqR9VYturX_10qMHRxMi/view?usp=sharing The zip file includes all the training data, and challenge tests

We recommend followers start with our data, since reproducing the pkl file requires stanford nlp parser and aligners, which causes many trouble for new users.

One of the checkpoints are provided in the following link. The checkpoint is one of the saved models being averaged in our paper's results. https://drive.google.com/file/d/1H-4MThtQbqTPgpl2SZBwNw_jlbeYNIqj/view?usp=sharing

Please put unzip data file to the data folder and run train file and test file respectively.

@article{feng2022neuro,
  title={Neuro-symbolic Natural Logic with Introspective Revision for Natural Language Inference},
  author={Feng, Yufei and Yang, Xiaoyu and Zhu, Xiaodan and Greenspan, Michael},
  journal={Transactions of the Association for Computational Linguistics},
  volume={10},
  pages={240--256},
  year={2022}
}