Large Language Models Have Unreliable Understanding of Software Engineering Terminology

2026-07-07

AI Summary

Large language models (LLMs) are increasingly employed in software engineering tasks, yet their reliability in understanding standardized terminology remains unclear. This study presents the first systematic evaluation of mainstream LLMs’ ability to discriminate definitions from the ISO/IEC/IEEE 24765:2017 standard. By constructing distractors through semantic substitution and structural deletion, and leveraging prompt engineering, the work assesses model performance in classification and reasoning. Findings reveal a prevalent “rejection-of-true” bias: while models accurately identify incorrect definitions, they frequently and erroneously reject correct ones. Moreover, explicit reasoning prompts not only fail to consistently improve performance but often degrade it due to overthinking. These results indicate that current LLMs lack robust comprehension of software engineering terminology as defined in established standards.

Abstract

Large Language Models (LLMs) are increasingly used in software engineering (SE), yet there is no systematic study that determines to which degree these LLMs actually understand standardized SE terminology. Lack of such understanding can lead to miscommunication and misunderstanding, both by LLMs consuming text but also by human-developers acting on LLM-generated text. Within this paper, we investigate to which degree state-of-the-art LLMs are able to identify whether definitions from the ISO/IEC/IEEE 24765:2017 Systems and Software Engineering - Vocabulary are correct. We prompt LLMs both with correct definitions, as well as systematically falsified definitions. The falsifications are both semantic (substitution of key terms) and structural (removing critical information). We measure both classification accuracy and whether reasoning tokens generated by the LLMs make sense with respect to understanding the definition. While most LLMs detect falsified definitions with high accuracy, they also reject many correct definitions, indicating a systematic rejection bias rather than genuine discriminative understanding. Explicit reasoning does not consistently improve results and may even hinder performance through over-thinking. Our work demonstrates that while the performance of LLMs (including their agentic use) in many SE tasks is impressive, there are still fundamental issues to understand how this will impact SE, including the consistent use of terminology.

Problem

• Large Language Models
• Software Engineering Terminology
• Terminology Understanding
• Miscommunication
• ISO/IEC/IEEE 24765

Innovation

• Large Language Models
• Software Engineering Terminology
• Terminology Understanding
• Definition Validation
• Reasoning Bias