This repository provides the benchmark dataset in paper NegotiationToM: A Benchmark for Stress-testing Machine Theory of Mind on Negotiation Surrounding.
The NegotiationToM dataset is in NegotiationToM.rar file. To prevent the data contamination issue, we set the password for the json file and the dataSet password is "NegotiationToM".
The details of this dataset and LLM performance are described in the following paper.
@article{DBLP:journals/corr/abs-2404-13627,
author = {Chunkit Chan and
Cheng Jiayang and
Yauwai Yim and
Zheye Deng and
Wei Fan and
Haoran Li and
Xin Liu and
Hongming Zhang and
Weiqi Wang and
Yangqiu Song},
title = {NegotiationToM: {A} Benchmark for Stress-testing Machine Theory of
Mind on Negotiation Surrounding},
journal = {CoRR},
volume = {abs/2404.13627},
year = {2024},
url = {https://doi.org/10.48550/arXiv.2404.13627},
doi = {10.48550/ARXIV.2404.13627},
eprinttype = {arXiv},
eprint = {2404.13627},
timestamp = {Wed, 26 Jun 2024 15:02:52 0200},
biburl = {https://dblp.org/rec/journals/corr/abs-2404-13627.bib},
bibsource = {dblp computer science bibliography, https://dblp.org}
}