Multilingual Prompting for Improving LLM Generation Diversity

TL;DR

This paper introduces multilingual prompting to enhance diversity in LLM outputs by incorporating cultural and linguistic cues, outperforming existing methods across various models and languages.

Contribution

It proposes a novel multilingual prompting technique that leverages cultural cues to improve diversity and reduce hallucinations in LLM generation.

Findings

Multilingual prompting outperforms existing diversity techniques.

Effectiveness varies with language resource levels and model size.

Aligning prompts with cultural cues reduces hallucinations.

Abstract

Large Language Models (LLMs) are known to lack cultural representation and overall diversity in their generations, from expressing opinions to answering factual questions. To mitigate this problem, we propose multilingual prompting: a prompting method which generates several variations of a base prompt with added cultural and linguistic cues from several cultures, generates responses, and then combines the results. Building on evidence that LLMs have language-specific knowledge, multilingual prompting seeks to increase diversity by activating a broader range of cultural knowledge embedded in model training data. Through experiments across multiple models (GPT-4o, GPT-4o-mini, LLaMA 70B, and LLaMA 8B), we show that multilingual prompting consistently outperforms existing diversity-enhancing techniques such as high-temperature sampling, step-by-step recall, and persona prompting. Further analyses show that the benefits of multilingual prompting vary between high and low resource languages and across model sizes, and that aligning the prompting language with cultural cues reduces hallucination about culturally-specific information.