Narrative-of-Thought (NoT)

Codebase for "NARRATIVE-OF-THOUGHT: Improving Temporal Reasoning of Large Language Models via Recounted Narratives"

Introduction

This repository contains codes and datasets for paper "Narrative-of-Thought: Improving Temporal Reasoning of Large Language Models via Recounted Narratives" (In Findings of the Association for Computational Linguistics: EMNLP 2024).

Set up

Run the following commands to clone the repository.

$ git clone https://github.com/launchnlp/NoT.git

Before running our codes, please run the following script to have all dependencies set up.

$ bash requirements.sh

*Our codebased is built on top of Llama-Factory. However, you are welcome to reproduce our work without relying on Llama-Factory, as the core methodology lies in the Narrative-of-Thought, which can be implemented by creating an inference loop using Huggingface alone.

Data

See Data directory for more information about how to download and process datasets.

Experiments

See model directory for more information about how to run this repo.

For the usgae of propriertary LLMs (e.g., GPT-3.5/4), you may set up by yourselves.

Citation

Please cite our paper if you use our codes and/or processed data:

@inproceedings{zhang-etal-2024-narrative,
    title = "Narrative-of-Thought: Improving Temporal Reasoning of Large Language Models via Recounted Narratives",
    author = "Zhang, Xinliang Frederick  and
      Beauchamp, Nicholas  and
      Wang, Lu",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2024",
    month = nov,
    year = "2024",
    address = "Miami, Florida, USA",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2024.findings-emnlp.963",
    pages = "16507--16530",
}