Consistency Analysis of ChatGPT
This addresses reliability concerns for users of large language models in applications requiring logical reasoning, though it is incremental as it builds on existing skepticism about AI consistency.
The paper investigated the logical consistency of ChatGPT and GPT-4, finding that despite enhanced language understanding, they frequently fail to generate logically consistent predictions, with experiments showing that prompt design, few-shot learning, and larger models are unlikely to fully resolve this issue.
ChatGPT has gained a huge popularity since its introduction. Its positive aspects have been reported through many media platforms, and some analyses even showed that ChatGPT achieved a decent grade in professional exams, adding extra support to the claim that AI can now assist and even replace humans in industrial fields. Others, however, doubt its reliability and trustworthiness. This paper investigates the trustworthiness of ChatGPT and GPT-4 regarding logically consistent behaviour, focusing specifically on semantic consistency and the properties of negation, symmetric, and transitive consistency. Our findings suggest that while both models appear to show an enhanced language understanding and reasoning ability, they still frequently fall short of generating logically consistent predictions. We also ascertain via experiments that prompt designing, few-shot learning and employing larger large language models (LLMs) are unlikely to be the ultimate solution to resolve the inconsistency issue of LLMs.