Large Language Models are Good Multi-lingual Learners : When LLMs Meet Cross-lingual Prompts

TL;DR

This paper introduces MLPrompt, a multilingual prompting strategy that improves large language models' reasoning on complex tasks by translating and highlighting challenging rules, outperforming existing methods.

Contribution

The paper proposes MLPrompt, a novel multilingual prompting technique that enhances LLM reasoning on complex contexts and integrates auto-checking for structured data generation.

Findings

MLPrompt outperforms Chain of Thought, Tree of Thought, and Self-Consistency methods.

The framework effectively generates structured data in text-to-MIP and text-to-SQL tasks.

Experimental results demonstrate improved accuracy and reasoning capabilities.

Abstract

With the advent of Large Language Models (LLMs), generating rule-based data for real-world applications has become more accessible. Due to the inherent ambiguity of natural language and the complexity of rule sets, especially in long contexts, LLMs often struggle to follow all specified rules, frequently omitting at least one. To enhance the reasoning and understanding of LLMs on long and complex contexts, we propose a novel prompting strategy Multi-Lingual Prompt, namely MLPrompt, which automatically translates the error-prone rule that an LLM struggles to follow into another language, thus drawing greater attention to it. Experimental results on public datasets across various tasks have shown MLPrompt can outperform state-of-the-art prompting methods such as Chain of Thought, Tree of Thought, and Self-Consistency. Additionally, we introduce a framework integrating MLPrompt with an auto-checking mechanism for structured data generation, with a specific case study in text-to-MIP instances. Further, we extend the proposed framework for text-to-SQL to demonstrate its generation ability towards structured data synthesis.