Runzhong Wang - 汪润中
/ˌru:ndʒɔ:n ˈwɑːŋ/ Any approximations are acceptable.
runzhong AT mit DOT edu

I am a Postdoctoral Associate at MIT, working with Prof. Connor Coley. I develop machine learning and optimization methods for molecular discovery.

I received my PhD in Computer Science and Engineering from Shanghai Jiao Tong University (SJTU), advised by Prof. Xiaokang Yang and Prof. Junchi Yan.

My research develops machine learning and optimization methods for problems with complex structures and hard constraints. A central focus is identifying molecules from mass spectrometry data: inferring chemical structures from the fragments an instrument measures. This work connects advances in constrained learning, graph matching, and generative modeling with experimentally validated discoveries in metabolomics, environmental chemistry, and biosynthesis.

Looking ahead, I aim to build AI systems that identify molecules with reliable confidence, learn from unannotated measurements, and guide the experiments needed to resolve uncertain structures.

Research overview connecting constrained machine learning, learning-guided optimization, molecular inverse modeling, and experimentally validated discoveries in clinical studies, environmental science, synthetic chemistry, and bioengineering.

My research connects constrained machine learning and optimization with molecular identification and experimentally validated discovery.

Recent News

  • [2026/09] GLACIER, our model for mass spectrum prediction through fragment detection, was accepted at NeurIPS 2026. Congratulations to Rui-Xi!
  • [2026/07] MassSpecGym in the Wild received the Best Paper Award at the ICML GenBio Workshop. Congratulations to Hongxuan and the team!
  • [2026/05] FRIGID, our diffusion model for generating molecules from mass spectra, was accepted at ICML 2026. Congratulations to Monte and Hongxuan!
  • [2025/10] I was selected for the 2026 ACS Wiley Computers in Chemistry (COMP) Outstanding Postdoc Award.

Code and Software

ms-pred
  • Open-source models for predicting tandem mass spectra and supporting molecular identification.
  • Implementations of ICEBERG, MARASON, and SCARF.
  • A common framework for comparing these models with NEIMS, MassFormer, 3DMolMS, and GrAFF-MS.
pygmtools Downloads
  • A Python library bringing classical and neural graph matching methods into a common interface.
  • Supports GPU acceleration and integration with deep learning workflows.
  • Documentation and examples for research, teaching, and practical applications.
  • Install with pip install pygmtools.
ThinkMatch
  • Implementations and benchmarks for learning correspondences between graphs.
  • Implementations of our models: NGM(v2), GANN-MGM, PCA-GM, CIE, AFAT, and LinSAT.
  • Includes comparison methods such as GMN, BBGM, and COMMON.
Awesome ML4CO
  • A curated collection of papers on machine learning for combinatorial optimization.
  • Contributions of additional papers are welcome.

Publications

* Equal contribution; † Corresponding author.

Neural Spectral Prediction for Structure Annotation with Tandem Mass Spectrometry.
Runzhong Wang, Mrunali Manjrekar, Babak Mahjour, Julian Avila-Pacheco, Joules Provenzano, Erin Reynolds, Magdalena Lederbauer, Eivgeni Mashin, Samuel Goldman, Mingxun Wang, Jing-Ke Weng, Desirée L. Plata, Clary B. Clish, Connor W. Coley.
Nature Methods, accepted in principle.
Outstanding poster award at Machine Learning in Drug Discovery Symposium 2025, Broad Institute.
[preprint] [code] [video]

GLACIER: Rethinking Mass Spectrum Prediction as an Object Detection Problem.
Rui-Xi Wang, Runzhong Wang†, Connor W. Coley†.
NeurIPS 2026.
[preprint]

MassSpecGym in the Wild: Uncovering and Correcting Evaluation Pitfalls in AI-Driven Molecule Discovery.
Hongxuan Liu, Roman Bushuiev, Ivy Lightheart, Mrunali Manjrekar, Anton Bushuiev, Magdalena Lederbauer, Filip Jozefov, Yinkai Wang, Soha Hassoun, Josef Sivic, James Taylor, Runzhong Wang, David Healey, Tomáš Pluskal, Connor W. Coley.
ICML GenBio Workshop 2026. Best Paper Award.
[preprint]

FRIGID: Scaling Diffusion-Based Molecular Generation from Mass Spectra at Training and Inference Time.
Montgomery Bohde, Hongxuan Liu, Mrunali Manjrekar, Magdalena Lederbauer, Shuiwang Ji, Runzhong Wang†, Connor W. Coley†.
ICML 2026.
[preprint] [code]

MacroSimGNN: Efficient and Accurate Prediction of Macromolecule Pairwise Similarity via A Graph Neural Network.
Jiale Shi, Runzhong Wang, Nathan J. Rebello, Jiarui Lu, Bradley D. Olsen, Debra J. Audus.
Macromolecules 2026 (vol. 59, no. 4, pp. 1885-1900).
[paper]

Generation as Search Operator for Test-Time Scaling of Diffusion-Based Combinatorial Optimization.
Yang Li, Lvda Chen, Haonan Wang, Runzhong Wang, Junchi Yan.
NeurIPS 2025.
[paper]

Optimization of robotic liquid handling as a capacitated vehicle routing problem.
*Guangqi Wu, *Runzhong Wang, Connor W. Coley. (* equal contribution)
Digital Discovery 2025.
[paper] [code]

Neural Graph Matching Improves Retrieval Augmented Generation in Molecular Machine Learning.
*Runzhong Wang, *Rui-Xi Wang, Mrunali Manjrekar, Connor W. Coley. (* equal contribution)
ICML 2025.
[paper] [code]

DiffMS: Diffusion Generation of Molecules Conditioned on Mass Spectra.
Montgomery Bohde, Mrunali Manjrekar, Runzhong Wang, Shuiwang Ji, Connor W. Coley.
ICML 2025.
[paper] [code]

Batched Bayesian optimization with correlated candidate uncertainties.
Jenna Fromer, Runzhong Wang, Mrunali Manjrekar, Austin Tripp, José Miguel Hernández-Lobato, Connor W. Coley.
JCIM 2025.
[paper]

Unify ML4TSP: Drawing Methodological Principles for TSP and Beyond from Streamlined Design Space of Learning and Search.
Yang Li, Jiale Ma, Wenzheng Pan, Runzhong Wang, Haoyu Geng, Nianzu Yang, Junchi Yan.
ICLR 2025.
[paper] [code]

Unified and Generalizable Reinforcement Learning for Facility Location Problems on Graphs.
Wenxuan Guo, Runzhong Wang, Yanyan Xu, Yaohui Jin.
WWW 2025.
[paper] [code]

Schedule Optimization for Chemical Library Synthesis.
Qianxiang Ai, Fanwang Meng, Runzhong Wang, J. Cullen Klein, Alexander G. Godfrey, Connor W. Coley.
Digital Discovery 2024.
[paper]

Fast T2T: Optimization Consistency Speeds Up Diffusion-Based Training-to-Testing Solving for Combinatorial Optimization.
Yang Li, Jinpei Guo, Runzhong Wang, Hongyuan Zha, Junchi Yan.
NeurIPS 2024.
[paper]

Double-Ended Synthesis Planning with Goal-Constrained Bidirectional Search.
Kevin Yu, Jihye Roh, Ziang Li, Wenhao Gao, Runzhong Wang, Connor W. Coley.
NeurIPS 2024.
[paper] [code]

Benchmarking PtO and PnO Methods in the Predictive Combinatorial Optimization Regime.
Haoyu Geng, Hang Ruan, Runzhong Wang, Yang Li, Yang Wang, Lei Chen, Junchi Yan.
NeurIPS 2024 Datasets and Benchmarks Track.
[paper] [code]

LLMCO4MR: LLMs-aided Neural Combinatorial Solver for Ancient Manuscript Restoration from Fragments with Case Studies on Dunhuang.
Yuqing Zhang, Hangqi Li, Shengyu Zhang, Runzhong Wang, Baoyi He, Huaiyong Dou, Junchi Yan, Yongquan Zhang, Fei Wu.
ECCV 2024.
[paper]

正线性约束组合优化问题的非自回归学习求解 | Learning to Solve Combinatorial Optimization under Positive Linear Constraints via Non-Autoregressive Neural Networks.
汪润中, 郦洋, 严骏驰, 杨小康 | Runzhong Wang, Yang Li, Junchi Yan, Xiaokang Yang.
中国科学：信息科学 | SCIENTIA SINICA Informationis 2024.
[中文论文 | English version] [code]

Pygmtools: A Python Graph Matching Toolkit.
Runzhong Wang, Ziao Guo, Wenzheng Pan, Jiale Ma, Yikai Zhang, Nan Yang, Qi Liu, Longxuan Wei, Hanxue Zhang, Chang Liu, Zetian Jiang, Xiaokang Yang, Junchi Yan.
JMLR 2024 (vol. 25, no. 33, pp. 1-7).
[paper] [code] [documentation]

Rethinking Cross-Domain Sequential Recommendation under Open-World Assumptions.
Wujiang Xu, Qitian Wu, Runzhong Wang, Mingming Ha, Qiongxu Ma, Linxun Chen, Bing Han, Junchi Yan.
WWW 2024.
[paper]

From Matching to Mixing: Learning Graph Interpolation for SAT Instance Generation.
Xinyan Chen, Yang Li, Runzhong Wang, Junchi Yan.
ICLR 2024.
[paper]

GMTR: Graph Matching Transformers.
Jinpei Guo, Shaofeng Zhang, Runzhong Wang, Chang Liu, Junchi Yan.
ICASSP 2024.
[paper]

T2T: From Distribution Learning in Training to Gradient Search in Testing for Combinatorial Optimization.
Yang Li, Jinpei Guo, Runzhong Wang, Junchi Yan.
NeurIPS 2023.
[paper] [code]

GAL-VNE: Solving the VNE Problem with Global Reinforcement Learning and Local One-Shot Neural Prediction.
Haoyu Geng, Runzhong Wang, Fei Wu, Junchi Yan.
KDD 2023.
[paper] [code]

InstaBoost++: Visual Coherence Principles for Unified 2D/3D Instance Level Data Augmentation.
Jianhua Sun, Hao-Shu Fang, Yuxuan Li, Runzhong Wang, Minghao Gou, Cewu Lu.
IJCV 2023 (vol. 131, pp. 2665-2681).
[paper] [code]

LinSATNet: The Positive Linear Satisfiability Neural Networks.
Runzhong Wang, Yunhao Zhang, Ziao Guo, Tianyi Chen, Xiaokang Yang, Junchi Yan.
ICML 2023.
[paper] [code]
Note: the Arxiv version fixed a minor issue in Eq (11) with the original ICML'23 publication

Unsupervised Learning of Graph Matching with Mixture of Modes via Discrepancy Minimization.
Runzhong Wang, Junchi Yan, Xiaokang Yang.
TPAMI 2023 (vol. 45, no. 8, pp. 10500-10518).
[paper] [project page] [code]

Deep Learning of Partial Graph Matching via Differentiable Top-K.
*Runzhong Wang, *Ziao Guo, Shaofei Jiang, Xiaokang Yang, Junchi Yan. (* equal contribution)
CVPR 2023.
[paper] [code]

MHSCNet: A Multimodal Hierarchical Shot-aware Convolutional Network for Video Summarization.
Wujiang Xu, Runzhong Wang, Xiaobo Guo, Shaoshuai Li, Qiongxu Ma, Yunan Zhao, Sheng Guo, Zhenfeng Zhu, Junchi Yan.
ICASSP 2023.
[paper]

Towards One-Shot Neural Combinatorial Optimization Solvers: Theoretical and Empirical Notes on the Cardinality-Constrained Case.
Runzhong Wang, Li Shen, Yiting Chen, Xiaokang Yang, Dacheng Tao, Junchi Yan.
ICLR 2023.
[paper] [code]

Revocable Deep Reinforcement Learning with Affinity Regularization for Outlier-Robust Graph Matching.
Chang Liu, Zetian Jiang, Runzhong Wang, Lingxiao Huang, Pinyan Lu, Junchi Yan.
ICLR 2023.
[paper] [code]

ROCO: A General Framework for Evaluating Robustness of Combinatorial Optimization Solvers on Graphs.
Han Lu, Zenan Li, Runzhong Wang, Qibing Ren, Xijun Li, Mingxuan Yuan, Jia Zeng, Xiaokang Yang, Junchi Yan.
ICLR 2023.
[paper] [code]

Deep Neural Network Fusion via Graph Matching with Applications to Model Ensemble and Federated Learning.
Chang Liu, Chenfei Lou, Runzhong Wang, Alan Yuhan Xi, Li Shen, Junchi Yan.
ICML 2022.
[paper] [code]

Appearance and structure aware robust deep visual graph matching: Attack, defense and beyond.
Qibing Ren, Qingquan Bao, Runzhong Wang, Junchi Yan.
CVPR 2022.
[paper] [code]

A Bi-Level Framework for Learning to Solve Combinatorial Optimization on Graphs.
Runzhong Wang, Zhigang Hua, Gan Liu, Jiayi Zhang, Junchi Yan, Feng Qi, Shuang Yang, Jun Zhou, Xiaokang Yang.
NeurIPS 2021.
[paper] [code]

Deep Latent Graph Matching.
Tianshu Yu, Runzhong Wang, Junchi Yan, Baoxin Li.
ICML 2021.
[paper]

Neural Graph Matching Network: Learning Lawler's Quadratic Assignment Problem with Extension to Hypergraph and Multiple-graph Matching.
Runzhong Wang, Junchi Yan, Xiaokang Yang.
TPAMI 2022 (vol. 44, no. 9, pp. 5261-5279).
Best poster award at International Conference on Data Science 2019.
[paper] [project page] [code]

Combinatorial Learning of Graph Edit Distance via Dynamic Embedding.
Runzhong Wang, Tianqi Zhang, Tianshu Yu, Junchi Yan, Xiaokang Yang.
CVPR 2021.
[paper] [code]

Graduated Assignment for Joint Multi-Graph Matching and Clustering with Application to Unsupervised Graph Matching Network Learning.
Runzhong Wang, Junchi Yan, Xiaokang Yang.
NeurIPS 2020.
[paper] [project page] [code]

Combinatorial Learning of Robust Deep Graph Matching: an Embedding based Approach.
Runzhong Wang, Junchi Yan, Xiaokang Yang.
TPAMI 2023 (vol. 45, no. 6, pp. 6984 - 7000).
[paper] [project page] [code]

Learning deep graph matching with channel-independent embedding and Hungarian attention.
Tianshu Yu, Runzhong Wang, Junchi Yan, Baoxin Li.
ICLR 2020.
[paper] [code]

Learning Combinatorial Embedding Networks for Deep Graph Matching.
Runzhong Wang, Junchi Yan, Xiaokang Yang.
ICCV 2019 (oral).
[paper] [project page] [code] [slides]

InstaBoost: Boosting Instance Segmentation via Probability Map Guided Copy-Pasting.
*Runzhong Wang, *Hao-shu Fang, *Jianhua Sun, Minghao Gou, Yong-Lu Li, Cewu Lu. (* equal contribution)
ICCV 2019.
[paper] [code]

Teaching and Mentoring

I have taught and mentored students across machine learning, optimization, and molecular discovery. At MIT, I guest lectured in Matrix Methods in Data Analysis, Signal Processing, and Machine Learning (18.065, Spring 2024) and completed the Kaufman Teaching Certificate Program. During my PhD at SJTU, I served as a teaching assistant for Deep Learning and Its Applications.

I have mentored 22 undergraduate and graduate researchers whose collective work has led to 23 papers. Student-led projects include DiffMS, FRIGID, and GLACIER; I served as co-corresponding author on FRIGID and GLACIER. My teaching interests span machine learning, optimization, and molecular machine learning, connecting foundational principles with scientific problems.

Doctoral Dissertation

排列型组合优化问题的机器学习求解方法研究.
上海交通大学学位论文.
Machine Learning Solvers for Permutation-based Combinatorial Optimization Problems.
A Dissertation Submitted to Shanghai Jiao Tong University for Doctoral Degree.
Runzhong Wang, Advisors: Xiaokang Yang, Junchi Yan.
[thesis (in Chinese)]

Work

2023 to present
Postdoctoral Associate
Department of Chemical Engineering and Department of EECS, MIT

Education

2019 to 2023
PhD in Computer Science and Engineering (with honors)
Wen-Tsün Wu Honorary Class, AI Institute
School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University (SJTU)

2015 to 2019
Bachelor of Engineering in Information Engineering
School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University (SJTU)

Professional Service

I review for machine learning and computer vision conferences, including NeurIPS, ICML, ICLR, CVPR, and ICCV, and journals including Nature Portfolio journals, TPAMI, and IJCV.