# ModGNN: Expert Policy Approximation in Multi-Agent Systems with a Modular Graph Neural Network Architecture

arXiv: https://arxiv.org/abs/2103.13446 · HTML: https://mlsanity.com/paper/2103.13446/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- Builds on 16 of its 20 references that are in the map.
- First posted on arXiv 24 Mar 2021.

## Builds on (most cited first, 16 of 16)

- [Semi-Supervised Classification with Graph Convolutional Networks](https://mlsanity.com/paper/1609.02907/) (2016)
- [Graph Attention Networks](https://mlsanity.com/paper/1710.10903/) (2017)
- [Inductive Representation Learning on Large Graphs](https://mlsanity.com/paper/1706.02216/) (2017)
- [Neural Message Passing for Quantum Chemistry](https://mlsanity.com/paper/1704.01212/) (2017)
- [Deep Sets](https://mlsanity.com/paper/1703.06114/) (2017)
- [Relational inductive biases, deep learning, and graph networks](https://mlsanity.com/paper/1806.01261/) (2018)
- [Link Prediction Based on Graph Neural Networks](https://mlsanity.com/paper/1802.09691/) (2018)
- [Convolutional Neural Network Architectures for Signals Supported on Graphs](https://mlsanity.com/paper/1805.00165/) (2018)
- [Graph Neural Networks for Decentralized Multi-Robot Path Planning](https://mlsanity.com/paper/1912.06095/) (2019)
- [Graph Neural Networks: Architectures, Stability and Transferability](https://mlsanity.com/paper/2008.01767/) (2020)
- [Learning Decentralized Controllers for Robot Swarms with Graph Neural Networks](https://mlsanity.com/paper/1903.10527/) (2019)
- [Graph Policy Gradients for Large Scale Robot Control](https://mlsanity.com/paper/1907.03822/) (2019)
- [Multi-Robot Coverage and Exploration using Spatial Graph Neural Networks](https://mlsanity.com/paper/2011.01119/) (2020)
- [Contextual Graph Markov Model: A Deep and Generative Approach to Graph Processing](https://mlsanity.com/paper/1805.10636/) (2018)
- [Graph Neural Networks for Learning Robot Team Coordination](https://mlsanity.com/paper/1805.03737/) (2018)
- [Graph Policy Gradients for Large Scale Unlabeled Motion Planning with Constraints](https://mlsanity.com/paper/1909.10704/) (2019)

## Built on by (most cited first, 9 of 9)

- [Multi-Robot Collaborative Perception with Graph Neural Networks](https://mlsanity.com/paper/2201.01760/) (2022)
- [Beyond Robustness: A Taxonomy of Approaches towards Resilient Multi-Robot Systems](https://mlsanity.com/paper/2109.12343/) (2021)
- [A Framework for Real-World Multi-Robot Systems Running Decentralized GNN-Based Policies](https://mlsanity.com/paper/2111.01777/) (2021)
- [Environment Optimization for Multi-Agent Navigation](https://mlsanity.com/paper/2209.11279/) (2022)
- [Graph learning in robotics: a survey](https://mlsanity.com/paper/2310.04294/) (2023)
- [The Holy Grail of Multi-Robot Planning: Learning to Generate Online-Scalable Solutions from Offline-Optimal Experts](https://mlsanity.com/paper/2107.12254/) (2021)
- [Permutation-Invariant Set Autoencoders with Fixed-Size Embeddings for Multi-Agent Learning](https://mlsanity.com/paper/2302.12826/) (2023)
- [LPAC: Learnable Perception-Action-Communication Loops with Applications to Coverage Control](https://mlsanity.com/paper/2401.04855/) (2024)
- [Learning Decentralized Flocking Controllers with Spatio-Temporal Graph Neural Network](https://mlsanity.com/paper/2309.17437/) (2023)

## Abstract (arXiv)

Recent work in the multi-agent domain has shown the promise of Graph Neural Networks (GNNs) to learn complex coordination strategies. However, most current approaches use minor variants of a Graph Convolutional Network (GCN), which applies a convolution to the communication graph formed by the multi-agent system. In this paper, we investigate whether the performance and generalization of GCNs can be improved upon. We introduce ModGNN, a decentralized framework which serves as a generalization of GCNs, providing more flexibility. To test our hypothesis, we evaluate an implementation of ModGNN against several baselines in the multi-agent flocking problem. We perform an ablation analysis to show that the most important component of our framework is one that does not exist in a GCN. By varying the number of agents, we also demonstrate that an application-agnostic implementation of ModGNN possesses an improved ability to generalize to new environments.