Summary of Collaborative Gym: a Framework For Enabling and Evaluating Human-agent Collaboration, by Yijia Shao et al.
Collaborative Gym: A Framework for Enabling and Evaluating Human-Agent Collaboration
by Yijia Shao, Vinay Samuel, Yucheng Jiang, John Yang, Diyi Yang
First submitted to arxiv on: 20 Dec 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: Computation and Language (cs.CL); Human-Computer Interaction (cs.HC)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This research paper presents a framework for studying human-agent collaboration, called Collaborative Gym (Co-Gym). Co-Gym enables asynchronous interaction among agents, humans, and task environments. The authors instantiate Co-Gym with three tasks in simulated and real-world conditions and propose an evaluation framework that assesses both the collaboration outcomes and processes. The results show that collaborative agents outperform their fully autonomous counterparts in task performance, achieving high win rates in various domains. However, the study highlights significant challenges in developing collaborative agents, requiring advancements in communication capabilities, situational awareness, and balancing autonomy and human control. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper creates a special platform called Co-Gym to help people understand how humans and machines work together. They test this platform with three different tasks that are like games, one for planning trips, one for analyzing data, and one for finding relevant information. The results show that when humans and machines work together, they can do better than if the machine worked alone. But the study also shows that making this happen is very hard because it requires machines to be able to talk and understand each other, know what’s going on around them, and figure out how much control humans should have. |
