Jiaxuan You (尤佳轩)

Ph.D. candidate in Computer Science
Stanford University
Email: jiaxuan@stanford.edu
Office: Gates 454
[Google scholar] [Github]

Hi! I am a 3rd year PhD student in Department of Computer Science, Stanford University. I'm advised by Prof. Jure Leskovec.
In the past, I have been fortunate enough to work with Prof. Stefano Ermon and Prof. David Lobell as a summer intern in 2017. In Tsinghua, I have worked on Machine Learning supervised by Prof. Jun Zhu.
My research focuses on developing Machine Learning algorithms for graph/relational structured data, including generating/optimizing graph structures, developing new graph neural network architectures, and applying these techniques in various domains.

Work Experience

  • Pinterest, Research Intern
    June 2018 - Sept 2018
    Mentor: Aditya Pal, Pong Eksombatchai, Chuck Rosenberg
    Developed large-scale dynamic recommender systems, published on WWW 2019.
  • Facebook AI Research (FAIR), Research Intern
    June 2019 - Feb 2020

    Mentor: Saining Xie, Kaiming He
    Graph Inspired Neural Network architecture design.

Professional Services

  • Program Committee member / Reviewer:
    Journals: IEEE Transactions on Pattern Analysis and Machine Intelligence, IEEE Access, IEEE Intelligent Systems, VLDB Journal
    Conferences: ICML 2019, SIGGRAPH 2019, NeurIPS 2019 (top-rated reviewer), AAAI 2020, WWW 2020, ICML 2020, ICWSM 2020
    Worshops: NeurIPS 2018 Workshop on Relational Representation Learning, ICLR 2019 Workshop on Representation Learning on Graphs and Manifolds, ICML 2019 Workshop on Learning and Reasoning with Graph-Structured Data, NeurIPS 2019 Workshop on Graph Representation Learning


  1. G2SAT: Learning to Generate SAT Formulas
    Jiaxuan You*, Haoze Wu*, Clark Barrett, Raghuram Ramanujan, Jure Leskovec.
    33th Conference on Neural Information Processing Systems (NeurIPS 2019)
    [PDF] [Code] [Webpage]
  2. GNNExplainer: A Tool for Post-hoc Explanation of Graph Neural Networks
    Rex Ying, Dylan Bourgeois, Jiaxuan You, Marinka Zitnik, Jure Leskovec
    33th Conference on Neural Information Processing Systems (NeurIPS 2019)
    [PDF] [Code] [Webpage]
  3. Position-aware Graph Neural Networks
    Jiaxuan You, Rex Ying, Jure Leskovec
    36th International Conference on Machine Learning (ICML 2019)
    Long Oral
    [PDF] [Code] [Webpage] [Video Recording]
  4. Redundancy-Free Computation Graphs for Graph Neural Networks​
    Zhihao Jia, Sina Lin, Rex Ying, Jiaxuan You, Jure Leskovec, Alex Aiken
    arXiv preprint
  5. Hierarchical Temporal Convolutional Networks for Dynamic Recommender Systems
    Jiaxuan You, Yichen Wang, Aditya Pal, Pong Eksombatchai, Chuck Rosenburg, Jure Leskovec
    The Web Conference 2019 (WWW 2019)
    [PDF] [Code]
  6. Graph Convolutional Policy Network for Goal-Directed Molecular Graph Generation
    Jiaxuan You*, Bowen Liu*, Rex Ying, Vijay Pande, Jure Leskovec
    32th Conference on Neural Information Processing Systems (NeurIPS 2018)
    Spotlight presentation
    [PDF] [Code]
  7. Hierarchical Graph Representation Learning with Differentiable Pooling
    Rex Ying, Jiaxuan You, Christopher Morris, Xiang Ren, William L. Hamilton, Jure Leskovec
    32th Conference on Neural Information Processing Systems (NeurIPS 2018)
    Spotlight presentation
    [PDF] [Code]
  8. GraphRNN: Generating Realistic Graphs with Deep Auto-regressive Model
    Jiaxuan You*, Rex Ying*, Xiang Ren, William L. Hamilton, Jure Leskovec
    35th International Conference on Machine Learning (ICML 2018)
    [PDF] [Code]
  9. Deep Gaussian Process for Crop Yield Prediction Based on Remote Sensing Data
    Jiaxuan You, Xiaocheng Li, Melvin Low, David Lobell, Stefano Ermon
    31th AAAI Conference on Artificial Intelligence (AAAI 2017)
    Oral, Best Student Paper Award (Computational Sustainability Track)
    [PDF] [Code] [Project Webpage]
  10. Scalable Crop Yield Prediction Approach by Combining Deep Learning with Remote Sensing Data
    Jiaxuan You, Xiaocheng Li, Stefano Ermon
    Best Big Data Solution in World Bank Big Data Innovation Challenge
    1st place among 180+ teams
    [link] [Supplementary Materials]
  11. An Effective Simulation Model for Multi-line Metro Systems Based on Origin-destination Data
    Jiaxuan You, Wei Guo, Yi Zhang, et al.
    19th IEEE International Conference on Intelligent Transportation Systems (ITSC 2016)
    As the only undergraduate attendee, I gave talks for 4 papers and was warmly welcomed
    [PDF] [Photo]
  12. Travel Modal Choice Analysis for Traffic Corridors Based on Decision-theoretic Approaches
    Wei Guo, Yi Zhang, Jiaxuan You, et al.
    Journal of Central South University (SCI, EI), Nov 2015.

Research Highlights

[Full image]

G2SAT: Learning to Generate SAT Formulas   (NeurIPS 2019)

The Boolean Satisfiability (SAT) problem is the canonical NP-complete problem that is fundamental to Computer Science. Developing and evaluating practical SAT solvers relies on extensive empirical testing on a set of real-world benchmark formulas. However, the number of such real-world SAT formulas is limited, and while augmenting benchmark formulas via synthetic SAT formulas is possible, existing approaches are heavily hand-crafted and cannot simultaneously capture a wide range of characteristics of real-world SAT formulas. Here we present G2SAT, the first deep generative framework that learns to generate SAT formulas from a given set of input formulas. The core of G2SAT is a novel deep generative model for bipartite graphs, which generates graphs via iterative node merging operations. We show that G2SAT can generate SAT formulas that closely resemble given real-world SAT formulas, as measured by both graph metrics and SAT solver behavior. Furthermore, we show that our synthetic SAT formulas can be used to improve SAT solver performance on real-world benchmarks, which opens up new opportunities for the continued development of SAT solvers and a deeper understanding of their performance.
[PDF] [Code] [Webpage]

GNNExplainer: A Tool for Post-hoc Explanation of Graph Neural Networks   (NeurIPS 2019)

GNNs combine node feature information with the graph structure by recursively passing neural messages along edges of the input graph. However, incorporating both graph structure and feature information leads to complex models and explaining predictions made by GNNs remains unsolved. Here we propose GNNExplainer, the first general, model-agnostic approach for providing interpretable explanations for predictions of any GNN-based model on any graph-based machine learning task. Given an instance, GNNExplainer identifies a compact subgraph structure and a small subset of node features that have a crucial role in GNN’s prediction. Further, GNNExplainer can generate consistent and concise explanations for an entire class of instances. Experiments on synthetic and real-world graphs show that our approach can identify important graph structures as well as node features, and outperforms baselines by 17.1% on average. GNNExplainer provides a variety of benefits, from the ability to visualize semantically relevant structures to interpretability, to giving insights into errors of faulty GNNs.
[PDF] [Code] [Webpage]

P-GNN: Position-aware Graph Neural Networks   (ICML 2019)

Learning node embeddings that capture a node’s position within the broader graph structure is crucial for many prediction tasks on graphs. However, existing Graph Neural Network (GNN) architectures have limited power in capturing the position/location of a given node with respect to all other nodes of the graph. Here we propose Position-aware Graph Neural Networks (PGNNs), a new class of GNNs for computing position-aware node embeddings. P-GNN first samples sets of anchor nodes, computes the distance of a given target node to each anchor-set, and then learns a non-linear distance-weighted aggregation scheme over the anchor-sets. This way P-GNNs can capture positions/locations of nodes with respect to the anchor nodes. We apply P-GNNs to multiple prediction tasks including link prediction and community detection. We show that P-GNNs consistently outperform state of the art GNNs, with up to 66% improvement in terms of the ROC AUC score.
[PDF] [Code] [Webpage] [Video Recording]

HAG: Redundancy-Free Computation Graphs for Graph Neural Networks​  

In submission.

HierTCN: Hierarchical Temporal Convolutional Networks for Dynamic Recommender Systems   (WWW 2019)

Existing dynamic recommender systems often use out-of-the-box sequence models which are limited by speed and memory consumption, are often infeasible for production environments, and usually do not incorporate cross-session information. Here we propose Hierarchical Temporal Convolutional Networks (HierTCN), a hierarchical deep learning architecture that makes dynamic recommendations based on users' sequential multi-session interactions with items. We conduct extensive experiments on a public XING dataset and a large-scale Pinterest dataset of 6 million users with 1.6 billion interactions. We show that HierTCN is 2.5x faster than RNN-based models and uses 90% less data memory compared to TCN-based models.
[PDF] [Code]

[Full image]

GCPN: Reinforcement Learning for Goal-Directed Molecular Graph Generation   (NeurIPS 2018)

Generating novel graph structures that optimize given objectives while obeying some given underlying rules is fundamental for chemistry, biology and social science research. Here we propose Graph Convolutional Policy Network (GCPN), a general graph convolutional network based model for goaldirected graph generation through reinforcement learning. The model is trained to optimize domain-specific rewards and adversarial loss through policy gradient, and acts in an environment that incorporates domain-specific rules. Experimental results show that GCPN can achieve 61% improvement on chemical property optimization over state-of-the-art baselines while resembling known molecules, and achieve 184% improvement on the constrained property optimization task.
[PDF] [Code

DiffPool: Differentiable Pooling layer for Graph Networks   (NeurIPS 2018)

Recently, graph neural networks (GNNs) have revolutionized the field of graph representation learning through effectively learned node embeddings. Here we propose DiffPool, a differentiable graph pooling module that can generate hierarchical representations of graphs and can be combined with various graph neural network architectures in an end-to-end fashion. Our experimental results show that combining existing GNN methods with DiffPool yields an average improvement of 5-10% accuracy on graph classification benchmarks, compared to all existing pooling approaches, achieving a new state-of-the-art on four out of five benchmark datasets.
[PDF] [Code]

GraphRNN: Generating Realistic Graphs with Deep Auto-regressive Model    (ICML 2018)

Modeling and generating graphs is fundamental for studying networks in biology, engineering, and social sciences. Here we propose GraphRNN, a deep autoregressive model that addresses the above challenges and approximates any distribution of graphs with minimal assumptions about their structure. In order to quantitatively evaluate the performance of GraphRNN, we introduce a benchmark suite of datasets, baselines and novel evaluation metrics based on Maximum Mean Discrepancy, which measure distances between sets of graphs. Our experiments show that GraphRNN significantly outperforms all baselines, learning to generate diverse graphs that match the structural characteristics of a target set, while also scaling to graphs 50× larger than previous deep models.
[PDF] [Code]

Crop Yield Prediction: Machine Learning over Satellite Images    (AAAI 2017)

Crop yield prediction is central in ensuring the food security. We introduce the first deep learning based method to predict crop yield purely based on publicly available remote sensing data. A novel dimensionality reduction method is proposed based on histogram calculation. By combining Gaussian Process with Convolutional Neural Networks, the model significantly outperforms traditional models, and we show that the features automatically extracted by deep learning models are much more effective than traditional hand-crafted features.
[PDF] [Code] [Project Webpage]

An Effective Simulation Model for Multi-line Metro Systems    (ITSC 2016)

Metro systems are playing an increasing role in modern cities, and various management solutions are proposed to improve their efficiency. When analyzing the effects of those solutions, simulation proves to be an effective and cost-saving method. This paper presents an effective simulation model for multi-line metro systems based on the OD (origin-destination) data and the network connection data. The model is validated in the scenario of Beijing metro system, which proves its effectiveness in large-scale empirical implementation.


  • I was born on Jan 5th, 1997.