Summary of Grasper: a Generalist Pursuer For Pursuit-evasion Problems, by Pengdeng Li et al.
Grasper: A Generalist Pursuer for Pursuit-Evasion Problems
by Pengdeng Li, Shuxin Li, Xinrun Wang, Jakub Cerny, Youzhi Zhang, Stephen McAleer, Hau Chan, Bo An
First submitted to arxiv on: 19 Apr 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: Computer Science and Game Theory (cs.GT); Multiagent Systems (cs.MA)
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 paper introduces Grasper, a novel architecture for solving pursuit-evasion games (PEGs) on graph-based environments. The authors focus on improving scalability by introducing a pre-training and fine-tuning paradigm using PSRO. Their contributions include a novel architecture combining graph neural networks and hypernetworks to generate pursuer policies, as well as an efficient three-stage training method involving self-supervised graph learning techniques like GraphMAE and heuristic-guided multi-task pre-training. The authors perform extensive experiments on synthetic and real-world maps, demonstrating Grasper’s superiority over baselines in terms of solution quality and generalizability. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Grasper is a new way to solve games where one team tries to catch another. This game is like navigating through an urban street network. The old methods didn’t work well when the starting conditions changed. Grasper can create good strategies for different game situations. It uses two main parts: a graph neural network and a hypernetwork. These work together to make good decisions. To train Grasper, the authors used a three-stage process that includes learning from examples, using general rules, and fine-tuning the strategy. The results show that Grasper does better than other methods in solving these games. |
Keywords
» Artificial intelligence » Fine tuning » Graph neural network » Multi task » Self supervised
