Mareike Hartmann wins Best Resource Paper Award at ACL 2024!

The paper “AppWorld: A Controllable World of Apps and People for Benchmarking Interactive Coding Agents” consists of an in-depth simulation of nine day-to-day web apps, such as Gmail and Spotify, along with complex natural-language tasks that can be solved by using these web apps. AppWorld demonstrates that such complex tasks are still hard to solve for current LLMs and offers an evaluation environment for developing new methods for tool use. Congratulations, Mareike!