Date
Publisher
arXiv
Large language models (LLMs) have shown increasing promise in educational settings, yet their mathematical reasoning has been considered evolving. This study evaluates the mathematical capabilities of various LLMs using the Finnish matriculation examination, a high-stakes digital test for upper secondary education. Initial tests yielded moderate performance corresponding to mid-range grades, but later evaluations demonstrated substantial improvements as the language models evolved. Remarkably, some models achieved near-perfect or perfect scores, matching top student performance and qualifying for university admission. Our findings highlight the rapid advances in the mathematical proficiency of LLMs and illustrate their potential to also support educational assessments at scale.
Study at a glance
- Question: How well can AI models solve math problems on a high-stakes high school exam?
- Setting and data: Finnish matriculation examination, a high-stakes digital test for upper secondary education.
- Method: Testing the performance of various language models on a digital mathematics examination.
Why it matters
Large language models have shown increasing promise in educational settings, yet their mathematical reasoning has been considered evolving.
Key findings
- On Finnish high-stakes upper secondary mathematics exams, some models achieved near-perfect or perfect scores, matching top students and qualifying for university admission.
- Earlier tests of the models yielded moderate mathematical performance corresponding to mid-range student grades.
- Subsequent evaluations demonstrated substantial improvements in the models' mathematical proficiency as the technology evolved.
Summary drafted with AI from the paper’s abstract and may contain errors. See the full paper for details.
Who age?
Why use AI?
Study design
