https://github.com/xf-zhao/LoT

Official implementation of LoT paper: "Enhancing Zero-Shot Chain-of-Thought Reasoning in Large Language Models through Logic"
https://github.com/xf-zhao/LoT

chain-of-thought chain-of-thought-prompting chain-of-thought-reasoning chain-of-thoughts coling coling2024 language language-model llm llms logic lot reasoning symbolic-logic symbolic-logic-reasoning zero-shot

Last synced: 3 months ago
JSON representation

Official implementation of LoT paper: "Enhancing Zero-Shot Chain-of-Thought Reasoning in Large Language Models through Logic"

• Host: GitHub
• URL: https://github.com/xf-zhao/LoT
• Owner: xf-zhao
• License: apache-2.0
• Created: 2023-06-10T10:57:02.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2024-03-13T12:20:41.000Z (11 months ago)
• Last Synced: 2024-08-01T02:27:29.989Z (6 months ago)
• Topics: chain-of-thought, chain-of-thought-prompting, chain-of-thought-reasoning, chain-of-thoughts, coling, coling2024, language, language-model, llm, llms, logic, lot, reasoning, symbolic-logic, symbolic-logic-reasoning, zero-shot
• Language: Python
• Homepage: https://arxiv.org/abs/2309.13339
• Size: 1.82 MB
• Stars: 10
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • Changelog: changelog.md
  • License: LICENSE

Awesome Lists containing this project

• Awesome-Reasoning-Foundation-Models - [Code

README

        
This is the repertory for the **LoT** paper:

 **"Enhancing Zero-Shot Chain-of-Thought Reasoning in Large Language Models through Logic"**, *[Xufeng Zhao](https://xf-zhao.github.io/), [Mengdi Li](https://mengdi-li.github.io/), [Wenhao Lu](https://lukaswill.github.io/), [Cornelius Weber](https://www.inf.uni-hamburg.de/en/inst/ab/wtm/people/weber.html), [Jae Hee Lee](https://jaeheelee.gitlab.io/), [Kun Chu](https://kchu.github.io/), [Stefan Wermter](https://www.inf.uni-hamburg.de/en/inst/ab/wtm/people/wermter.html).* [[arXiv](https://arxiv.org/abs/2309.13339)], to appear at `COLING 2024`.

![](https://github.com/xf-zhao/xf-zhao.github.io/blob/master/assets/img/publication_preview/LoT%20prompting.gif?raw=true)

Abstract:

Recent advancements in large language models have showcased their remarkable generalizability across various domains. However, their reasoning abilities still have significant room for improvement, especially when confronted with scenarios requiring multi-step reasoning. Although large language models possess extensive knowledge, their reasoning often fails to effectively utilize this knowledge to establish a coherent thinking paradigm. These models sometimes show hallucinations as their reasoning procedures are unconstrained by logical principles. Aiming at improving the zero-shot chain-of-thought reasoning ability of large language models, we propose LoT (Logical Thoughts), a self-improvement prompting framework that leverages principles rooted in symbolic logic, particularly Reductio ad Absurdum, to systematically verify and rectify the reasoning processes step by step. Experimental evaluations conducted on language tasks in diverse domains, including arithmetic, commonsense, symbolic, causal inference, and social problems, demonstrate the efficacy of enhanced reasoning by logic.

# Install

```bash

git clone https://github.com/xf-zhao/LoT.git

cd LoT

pip install -r requirements.txt

```

# Quick Test Run

```bash

python main.py --env cot --model_name gpt-3.5-turbo

python main.py --env lot --model_name gpt-3.5-turbo

```

# Run Exepriments

```bash

# vanilla CoT

sh run_cot.sh

# LoT with modified CoT, specifically, with "#1." as additional prompt

sh run_lot.sh

```

# Dataset

GSM8K: 

# Cite with BibTex

```text

@inproceedings{Zhao23EnhancingZeroShot,

  title = {Enhancing {{Zero-Shot Chain-of-Thought Reasoning}} in {{Large Language Models}} through {{Logic}}},

  author = {Zhao, Xufeng and Li, Mengdi and Lu, Wenhao and Weber, Cornelius and Lee, Jae Hee and Chu, Kun and Wermter, Stefan},

  booktitle = {2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024), Turin, Italy},

  year = {2024},

  month = may,

  keywords = {Large Language Models, Chain-of-Thought, Reasoning, Logic, Reductio ad Absurdum},

}

```