(Why) Is My Prompt Getting Worse? Rethinking Regression Testing for Evolving LLM APIs
This addresses the issue for downstream application developers who rely on LLM APIs, but it is incremental as it builds on existing testing concepts.
The paper tackles the problem of performance regression and prompt design instability caused by silent updates and deprecations of LLM APIs, as shown in a case study on toxicity detection. It emphasizes the need to rethink regression testing for these evolving APIs due to challenges like different correctness notions and non-determinism.
Large Language Models (LLMs) are increasingly integrated into software applications. Downstream application developers often access LLMs through APIs provided as a service. However, LLM APIs are often updated silently and scheduled to be deprecated, forcing users to continuously adapt to evolving models. This can cause performance regression and affect prompt design choices, as evidenced by our case study on toxicity detection. Based on our case study, we emphasize the need for and re-examine the concept of regression testing for evolving LLM APIs. We argue that regression testing LLMs requires fundamental changes to traditional testing approaches, due to different correctness notions, prompting brittleness, and non-determinism in LLM APIs.