Evaluating Multimodal Generative AI with Korean Educational Standards

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Park, Sanghee, Kim, Geewook
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910838818865152
author Park, Sanghee
Kim, Geewook
author_facet Park, Sanghee
Kim, Geewook
contents This paper presents the Korean National Educational Test Benchmark (KoNET), a new benchmark designed to evaluate Multimodal Generative AI Systems using Korean national educational tests. KoNET comprises four exams: the Korean Elementary General Educational Development Test (KoEGED), Middle (KoMGED), High (KoHGED), and College Scholastic Ability Test (KoCSAT). These exams are renowned for their rigorous standards and diverse questions, facilitating a comprehensive analysis of AI performance across different educational levels. By focusing on Korean, KoNET provides insights into model performance in less-explored languages. We assess a range of models - open-source, open-access, and closed APIs - by examining difficulties, subject diversity, and human error rates. The code and dataset builder will be made fully open-sourced at https://github.com/naver-ai/KoNET.
format Preprint
id arxiv_https___arxiv_org_abs_2502_15422
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Evaluating Multimodal Generative AI with Korean Educational Standards
Park, Sanghee
Kim, Geewook
Computation and Language
Artificial Intelligence
Computer Vision and Pattern Recognition
This paper presents the Korean National Educational Test Benchmark (KoNET), a new benchmark designed to evaluate Multimodal Generative AI Systems using Korean national educational tests. KoNET comprises four exams: the Korean Elementary General Educational Development Test (KoEGED), Middle (KoMGED), High (KoHGED), and College Scholastic Ability Test (KoCSAT). These exams are renowned for their rigorous standards and diverse questions, facilitating a comprehensive analysis of AI performance across different educational levels. By focusing on Korean, KoNET provides insights into model performance in less-explored languages. We assess a range of models - open-source, open-access, and closed APIs - by examining difficulties, subject diversity, and human error rates. The code and dataset builder will be made fully open-sourced at https://github.com/naver-ai/KoNET.
title Evaluating Multimodal Generative AI with Korean Educational Standards
topic Computation and Language
Artificial Intelligence
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2502.15422