Recent years have seen substantial progress in neural generation of text, images, and audio, supported by mature training pipelines and large-scale optimization. For graphs, however, comparable progress has been more limited. We attribute this gap to graph-specific optimization and representation challenges that undermine the effectiveness of training neural networks with backpropagation and gradient descent. We argue that representing graphs on a fixed-size Euclidean grid, as is common in recently proposed models for supervised graph prediction, may not be the optimal choice in these settings. To support our view, we provide an analysis of neural graph generation methods and identify theoretical challenges that lead to pitfalls when training neural networks to produce graphs as their output. Motivated by this analysis, we introduce \textbf{GRA}ph~\textbf{I}mitation~\textbf{L}earning~(GRAIL), a framework for training neural networks in supervised settings in which the supervision signal is a graph. GRAIL generates graphs sequentially through a Markov decision process over embeddings of partial graphs, thereby avoiding the representation issues associated with fixed-size grid graph representations. We empirically show that GRAIL achieves competitive results on supervised graph prediction across a comprehensive suite of 18 benchmarks, matching or surpassing state-of-the-art methods in several settings.

翻译：近年来，在文本、图像和音频的神经网络生成方面取得了显著进展，这得益于成熟的训练流程和大规模优化。然而，对于图而言，类似的进展较为有限。我们将这一差距归因于图特有的优化和表征挑战，这些挑战削弱了使用反向传播和梯度下降训练神经网络的有效性。我们认为，将图表示为固定大小的欧几里得网格（这在近期提出的有监督图预测模型中很常见）可能并非这些场景中的最优选择。为支持这一观点，我们分析了神经图生成方法，并识别出在训练神经网络以输出图时导致陷阱的理论挑战。受此分析启发，我们引入了\textbf{GRA}ph~\textbf{I}mitation~\textbf{L}earning（GRAIL）框架，用于在监督信号为图的有监督场景中训练神经网络。GRAIL通过基于部分图嵌入的马尔可夫决策过程顺序生成图，从而避免了固定大小网格图表征带来的表征问题。我们在包含18个基准测试的综合套件上进行了实证研究，结果表明GRAIL在有监督图预测任务中取得了具有竞争力的结果，在多个场景中匹配或超越了现有最先进方法。