Investigating Hallucination in Conversations for Low Resource Languages

TL;DR

This paper investigates hallucination issues in conversational LLMs across Hindi, Farsi, and Mandarin, revealing language-dependent differences in factual accuracy and analyzing multiple models' performance.

Contribution

It extends hallucination analysis to low-resource languages in conversational settings, providing a comprehensive dataset and comparison across several LLMs.

Findings

Fewer hallucinations in Mandarin responses

Higher hallucination rates in Hindi and Farsi

Model performance varies significantly by language

Abstract

Large Language Models (LLMs) have demonstrated remarkable proficiency in generating text that closely resemble human writing. However, they often generate factually incorrect statements, a problem typically referred to as 'hallucination'. Addressing hallucination is crucial for enhancing the reliability and effectiveness of LLMs. While much research has focused on hallucinations in English, our study extends this investigation to conversational data in three languages: Hindi, Farsi, and Mandarin. We offer a comprehensive analysis of a dataset to examine both factual and linguistic errors in these languages for GPT-3.5, GPT-4o, Llama-3.1, Gemma-2.0, DeepSeek-R1 and Qwen-3. We found that LLMs produce very few hallucinated responses in Mandarin but generate a significantly higher number of hallucinations in Hindi and Farsi.