Large Language Models (LLMs) are being quickly integrated in a myriad of software applications. This may introduce a number of biases, such as gender, age or ethnicity, in the behavior of such applications. To face this challenge, we explore the automatic generation of tests to assess the potential biases of an LLM.
Clustering of model instances by using graph kernels. Make sure you test your models with the most diverse set of examples!
Quality aspects of an API (availability, performance,…) are key aspects to take into account when deciding which API to choose. Our testing framework provides some insights on these non-functional properties as they are typically not disclosed
We present our ecosystem of tools to facilitate the automatic discovery, merging, quality assurance and code-generation of REST APIs, relying on standard specifications like OpenAPI and OData.
We automatically test REST API based on their specifications, particularly OpenAPI ones. To our surprise, many of our tests failed, meaning that the OpenAPI definition and the actual API were not an exact match.