Large Language Models Only Pass Primary School Exams in Indonesia: A Comprehensive Test on IndoMMLU

TL;DR

This paper introduces IndoMMLU, a comprehensive Indonesian language understanding benchmark, revealing that GPT-3.5 and other models only pass primary school level in Indonesia and have limited local cultural knowledge.

Contribution

The creation of IndoMMLU as the first multi-task benchmark for Indonesian language and culture understanding, enabling better evaluation of multilingual LLMs beyond English.

Findings

GPT-3.5 passes primary school exams in Indonesia

Smaller models perform even worse

Limited knowledge of Indonesian languages and culture

Abstract

Although large language models (LLMs) are often pre-trained on large-scale multilingual texts, their reasoning abilities and real-world knowledge are mainly evaluated based on English datasets. Assessing LLM capabilities beyond English is increasingly vital but hindered due to the lack of suitable datasets. In this work, we introduce IndoMMLU, the first multi-task language understanding benchmark for Indonesian culture and languages, which consists of questions from primary school to university entrance exams in Indonesia. By employing professional teachers, we obtain 14,981 questions across 64 tasks and education levels, with 46% of the questions focusing on assessing proficiency in the Indonesian language and knowledge of nine local languages and cultures in Indonesia. Our empirical evaluations show that GPT-3.5 only manages to pass the Indonesian primary school level, with limited knowledge of local Indonesian languages and culture. Other smaller models such as BLOOMZ and Falcon perform at even lower levels.