About

Nan Jiang is an Associate Professor at the Siebel School of Computing and Data Science at the University of Illinois Urbana-Champaign. He holds a PhD in Computer Science and Engineering from the University of Michigan, obtained in 2017, and a Bachelor of Engineering in Automation from Tsinghua University, earned in 2011. His research interests focus on Reinforcement Learning within the broader areas of Artificial Intelligence. Jiang has contributed to the field through his work on reinforcement learning theory and algorithms, and he serves as an action editor for prominent journals such as Statistical Science and the Journal of Machine Learning Research, as well as an editor for Foundations and Trends in Machine Learning. He has received multiple honors, including the NSF CAREER Award, a Sloan Research Fellowship, and recognition for his teaching excellence. Jiang is actively involved in research, teaching, and editorial activities, and he is recognized for his contributions to advancing reinforcement learning and AI.

Research topics

• Computer Science
• Artificial Intelligence
• Computer Security
• Machine Learning
• Information Retrieval
• Internet privacy
• Human–computer interaction
• World Wide Web
• Embedded system
• Art history
• Computer vision

Selected publications

• Multi-Scale Computational Analysis of Wikipedia’s Telling of Global History

  Research Square · 2026-01-19

  preprintOpen access
  PublisherDOI

• Anchor-SAM: Active Mining of Latent Anchors From SAM Encoder for Road Extraction

  IEEE Transactions on Geoscience and Remote Sensing · 2026-01-01

  articleSenior author

  Road extraction in high-resolution remote sensing imagery remains a persistent challenge due to occlusions and complex backgrounds, which lead to fragmented road topologies. However, existing road extraction methods constrained by limited pre-training remote sensing data often lack the generalization capability to distinguish roads from complex backgrounds. To address this issue, we propose Anchor-SAM, a novel framework that actively mines latent semantic anchors embedded in the SAM encoder to guide topological reconstruction. Our approach stems from a pivotal insight: the SAM encoder is able to abstract complex scenes into sparse semantic anchors at deep layers, thereby implicitly encoding the global structural skeleton. To harness these implicit cues, we introduce the Multi-scale Deformable Context Perceiver (MDCP) and the Deformable Bayesian Conditional Interaction Module (DBCIM). The MDCP explicitly utilizes spatial cues to aggregate global semantics across distributed anchors, establishing a robust initial context for the decoder. The DBCIM facilitates the diffusion of semantic cues to surrounding regions and effectively suppresses noise. Specifically, by leveraging the semantic certainty of anchors to guide deformable sampling trajectories, this mechanism proactively filters out background regions while precisely repairing fragmented road topologies. Our method achieves competitive performance on both the DeepGlobe and Massachusetts datasets. The source code will be publicly available at https://github.com/Hmbb0606/Anchor-SAM.

  PublisherDOI

• MAJSCC: Mamba-based adaptive joint source channel coding for wireless image transmission

  Journal of Electronic Imaging · 2026-03-06

  articleCorresponding

  Lightweight and efficient neural network models for joint source-channel coding (JSCC) are critical for advancing semantic communication. We propose an adaptive JSCC architecture, named Mamba-based adaptive joint source-channel coding (MAJSCC), which is explicitly designed for wireless image transmission tasks. By integrating the Mamba architecture into both the encoder and decoder, the architecture enhances local feature representation and incorporates an adaptive mechanism to flexibly adjust to diverse channel conditions and transmission rates. Furthermore, the proposed network utilizes wavelet convolution to exploit a broader spectrum of signal information during training, thereby improving its ability to capture high-resolution image details. Comprehensive experimental evaluations demonstrate that the proposed MAJSCC achieves comparable or superior performance in large-scale, high-resolution image transmission tasks. Compared with the state-of-the-art BPG + 5G LDPC-coded systems (executed on CPU), it delivers faster end-to-end encoding speeds (accelerated on graphics processing unit), with a compact model design that ensures higher efficiency than traditional CNN-based JSCC methods.

  PublisherDOI

• SegEdit: image editing via semantic mask segmentation and shape injection within diffusion model

  Expert Systems with Applications · 2026-01-22

  articleCorresponding
  PublisherDOI

• Mechanical stress accelerates vascular calcification by Piezo1/BMP2 of vascular smooth muscle cells

  Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease · 2025-09-27

  article
  PublisherDOI

• Introduction to the Special Issue on Reinforcement Learning

  Statistical Science · 2025-11-01 · 1 citations

  article1st authorCorresponding
  PublisherDOI

• Reinforce-Ada: An Adaptive Sampling Framework under Non-linear RL Objectives

  arXiv (Cornell University) · 2025-10-06

  preprintOpen access

  Reinforcement learning (RL) for large language model reasoning is frequently hindered by signal loss, a phenomenon where standard uniform sampling with small group sizes fails to uncover informative learning signals for difficult prompts. We demonstrate that this collapse is a statistical artifact of undersampling rather than an inherent model limitation. To address this systematically, we introduce a theoretical framework based on optimizing a non-linear RL objective (e.g., log-likelihood). We show that this objective naturally induces a weighted gradient estimator that prioritizes difficult prompts, which can be robustly realized through adaptive sampling. Guided by this framework, we propose Reinforce-Ada, a family of algorithms that dynamically allocates inference budgets based on prompt difficulty, effectively scaling up RL compute to where it is needed most. Unlike passive filtering methods that discard low-signal prompts, Reinforce-Ada actively invests compute to recover them. We introduce two efficient realizations: an estimation-based approach and a model-free sequential sampling approach. Extensive experiments across multiple benchmarks show that Reinforce-Ada significantly outperforms uniform baselines like GRPO, recovering lost signals and accelerating convergence by up to $2\times$ while maintaining the same total inference budget. Code is available at https://github.com/RLHFlow/Reinforce-Ada.

  PublisherOA PDFDOI

• MLSTAM: A multidimensional long-term spatio-temporal attention model for traffic flow forecasting by capturing time series correlations

  Intelligent Data Analysis · 2025-05-14

  article

  Traffic flow prediction occupies a pivotal position in intelligent transportation systems, and accurate traffic flow prediction is of great significance for alleviating traffic congestion and reducing the incidence of traffic accidents. To improve the accuracy of traffic flow forecasts, it is necessary to consider the historical data over a longer period. However, most of the existing methods only consider part of the recent historical time information, ignoring the implied fluctuation of the traffic flow in some regions in the historical contemporaneous time interval. Therefore, we propose a multidimensional long-term spatio-temporal attention model for traffic flow forecasting by capturing time series correlations. In this model, we design a multi-temporal dimensional attention mechanism and a deep fusion extraction convolutional neural network to capture multidimensional temporal information and fuse spatio-temporal correlations to predict traffic flow. The experimental results on two real datasets show that the proposed model outperforms the compared models.

  PublisherDOI

• A multi-agent deep reinforcement learning model for crowd coordinated evacuation simulation in complex environments

  Simulation Modelling Practice and Theory · 2025-12-03

  article
  PublisherDOI

• Spatio-Temporal Heterogeneity-Oriented Graph Convolutional Network for Urban Traffic Flow Prediction

  Sensors · 2025-08-18 · 2 citations

  articleOpen accessSenior author

  In the realm of urban vehicular ad hoc networks (VANETs), cross-domain data has constituted a multifaceted amalgamation of information sources, which significantly enhances the accuracy and response speed of traffic prediction. However, the interplay between spatial and temporal heterogeneity will complicate the complexity of geographical locations or physical connections in the data normalization. Besides, the traffic pattern differences incurred by dynamic external factors also bring cumulative and sensitive impacts during the construction of the prediction model. In this work, we propose the spatio-temporal heterogeneity-oriented graph convolutional network (SHGCN) to tackle the above challenges. First, the SHGCN analytically employs spatial heterogeneity between urban streets rather than simple adjacency relationships to reveal the spatio-temporal correlations of traffic stream movement. Then, the air quality data is taken as external factors to identify the traffic forecasting trend at the street level. The hybrid model of the graph convolutional network (GCN) and gated recurrent unit (GRU) is designed to investigate cross-correlation characteristics. Finally, with the real-world urban datasets, experimental results demonstrate that the SHGCN achieves improvements, with the RMSE and MAE reductions ranging from 2.91% to 41.26% compared to baseline models. Ablation studies confirm that integrating air quality factors with traffic patterns enhances prediction performance at varying degrees, validating the method's effectiveness in capturing the complex correlations among air pollutants, traffic flow dynamics, and road network topology.

  PublisherOA PDFDOI

Frequent coauthors

• Honglong Chen

  China University of Petroleum, East China

  18 shared

• Alekh Agarwal

  16 shared

• Lingfeng Liu

  14 shared

• Tao Wan

  East China Jiaotong University

  13 shared

• Tengyang Xie

  12 shared

• Akshay Krishnamurthy

  11 shared

• William J. Dally

  Nvidia (United States)

  11 shared

• Yexiang Xue

  10 shared

Labs

• Siebel School of Computing and Data SciencePI

Awards & honors

• Google Research Scholar Award (03/2024)
• Sloan Research Fellowship (02/2024)
• ICML 2022 Outstanding Paper Runner Up (07/2022)
• NSF CAREER Award (03/2022)
• AAMAS 2015 Best Paper Award (05/2015)