MacBehaviour: An R package for behavioural experimentation on large language models

There has been increasing interest in investigating the behaviours of large language models (LLMs) and LLM-powered chatbots. This paper presents the development of an R package called "MacBehaviour" that aims to streamline the experimental process when interfacing with LLMs for behavioural research. "MacBehaviour" offers a user-friendly package for conducting behavioural experimentation and for norming experimental stimuli with LLMs in the R environment. The package provides a suite of functions tailored for experiments on LLMs, including the design of experiment, the presentation of stimuli, and the manipulation of model behaviour. The package interfaces with APIs of an array of LLMs, including those from OpenAI's GPT series, Llama-2-chat-hf series in Hugging Face, and open-source models. Overall, "MacBehaviour" serves as a valuable resource for researchers, offering a user-friendly interface and comprehensive tools to streamline the experimental process and advance the study of machine behavior within the field of psychology.