What Makes Math Word Problems Challenging for LLMs?

TL;DR

This paper analyzes linguistic and mathematical factors that make math word problems challenging for large language models, using classifiers to predict difficulty and improve understanding of LLM performance on different problem categories.

Contribution

It introduces a feature-based analysis of MWPs, identifying key factors affecting LLM difficulty and enhancing prediction of model performance across problem types.

Findings

Certain linguistic features increase problem difficulty for LLMs

Mathematical complexity correlates with lower LLM accuracy

Classifier-based predictions improve understanding of LLM strengths and weaknesses

Abstract

This paper investigates the question of what makes math word problems (MWPs) in English challenging for large language models (LLMs). We conduct an in-depth analysis of the key linguistic and mathematical characteristics of MWPs. In addition, we train feature-based classifiers to better understand the impact of each feature on the overall difficulty of MWPs for prominent LLMs and investigate whether this helps predict how well LLMs fare against specific categories of MWPs.