Summary of Human-agent Cooperation in Games Under Incomplete Information Through Natural Language Communication, by Shenghui Chen et al.
Human-Agent Cooperation in Games under Incomplete Information through Natural Language Communication
by Shenghui Chen, Daniel Fried, Ufuk Topcu
First submitted to arxiv on: 23 May 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: 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 The proposed shared-control game is designed to evaluate autonomous agents’ ability to strategize and cooperate with humans under incomplete information. The game involves two players taking alternating turns to achieve a common objective, with each player having limited knowledge about the other’s actions. To solve this policy synthesis problem, the authors propose a communication-based approach comprising a language module and a planning module. The language module translates natural language messages into a compact representation of flags, which are used by the planning module to compute a policy using an asymmetric information-set Monte Carlo tree search algorithm. The effectiveness of this approach is evaluated in a testbed based on Gnomes at Night, a search-and-find maze board game, through human subject experiments. Results show that communication narrows the information gap between players and enhances human-agent cooperation efficiency with fewer turns. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This study is about creating robots that can work together with humans to achieve a common goal when there’s not enough information to make decisions. The researchers designed a special game where two people take turns trying to find a hidden treasure, but they don’t know what the other person will do next. They created a way for the robot to understand human language and use that information to make better choices. In tests with real people, this approach worked well and helped humans and robots work together more efficiently. |
