1 paper
Alexandre Drouin, Maxime Gasse, Massimo Caccia +9
We study the use of large language model-based agents for interacting with software via web browsers. Unlike prior work, we focus on measuring the agents' ability to perform tasks…