Innovative Research Award

Jindong Wang — College of William & Mary, United States

Jindong Wang is a computer scientist whose research spans machine learning, transfer learning, large language models, foundation models, federated learning, trustworthy artificial intelligence, and evaluation. His reported scholarly record includes 199 documents, more than 28,000 citations, and an h-index of 61, alongside research outputs appearing in major venues in artificial intelligence and machine learning.

His academic profile combines theoretical and applied machine learning with open-source research infrastructure, interdisciplinary evaluation, and responsible AI. His work includes widely used resources for transfer learning, semi-supervised learning, robust machine learning, personalized federated learning, and large-language-model evaluation. The profile also records academic service, teaching, invited talks, awards, grants, and research collaborations across academia and industry.[1]

Abstract

Jindong Wang’s research profile reflects sustained work across machine learning foundations, transfer learning, large language models, generative AI, federated learning, and responsible AI. His reported record combines more than 100 research publications, open-source systems, scholarly service, teaching, and externally supported projects, with selected publications receiving substantial citation attention. [1]

Keywords

Machine learning; transfer learning; domain generalization; large language models; foundation models; generative AI; multimodal AI; semi-supervised learning; federated learning; trustworthy AI; responsible AI; robustness; generalization; AI evaluation; personalized AI.

Introduction

Jindong Wang’s research addresses methods for developing machine-learning systems that can generalize across domains, adapt to new data and tasks, and operate reliably under changing conditions. Earlier work emphasized transfer learning, domain generalization, federated learning, and semi-supervised learning, while more recent research extends these interests toward foundation models, large language models, multimodal systems, evaluation, personalization, and trustworthy AI.

His professional experience includes an assistant professorship at William & Mary beginning in 2025 and previous research leadership at Microsoft Research. His academic training includes a Ph.D. in Computer Science from the Institute of Computing Technology, Chinese Academy of Sciences, following undergraduate study in engineering at North China University of Technology. He also undertook visiting research at The Hong Kong University of Science and Technology.

Research Profile

The research profile covers AI foundations, large language models and generative AI, and responsible AI. In AI foundations, the work includes machine learning, transfer learning, out-of-distribution generalization, semi-supervised learning, and federated learning. In generative AI, the emphasis includes LLM understanding and evaluation, agentic systems, multimodal models, fine-tuning, and adaptation. Responsible AI interests include human-centered AI, AI and society, trustworthiness, and alignment.[1]

• AI Foundations: machine learning, transfer learning, OOD generalization, semi-supervised learning, and federated learning.
• Large Language Models and Generative AI: LLM evaluation, foundation-model understanding, agentic AI, multimodal models, fine-tuning, and adaptation.
• Responsible AI: human-centered AI, trustworthy machine learning, AI and society, safety, and alignment.

Professional service has included editorial responsibilities and program-committee leadership for journals and conferences such as IEEE Transactions on Neural Networks and Learning Systems, Journal of Computer Science and Technology, ACM Transactions on Intelligent Systems and Technology, NeurIPS, ICLR, ICML, KDD, IJCAI, AAAI, CIKM, and AIES.

Research Contributions

A central contribution of Jindong Wang’s work is the development and synthesis of methods for transfer learning and domain generalization, including surveys and algorithms addressing adaptation to previously unseen distributions. These contributions provide conceptual and methodological foundations for learning systems that must operate beyond their original training environments. [2]

His research also contributes to semi-supervised and federated learning. FlexMatch introduced curriculum pseudo-labeling for semi-supervised learning, while related work has examined federated transfer learning and healthcare-oriented distributed learning. Such studies address settings where labeled data, centralized training, or uniform data distributions are limited. [3] [4]

More recent contributions focus on large language models and foundation models, particularly their evaluation, robustness, personalization, multimodal capabilities, and societal implications. The research direction emphasizes systematic assessment rather than relying solely on aggregate benchmark scores, connecting model behavior with robustness, bias, generalization, and practical deployment considerations.

Open-source activities complement the publication record. The Transfer Learning repository, PromptBench, USB, RobustLearn, and PersonalizedFL provide research materials, implementations, datasets, evaluation utilities, or experimental infrastructure. Their reported adoption illustrates an emphasis on reproducibility and community-accessible research tooling.

Publications

Jindong Wang’s publications advance transfer learning, domain generalization, semi-supervised learning, federated learning, and large-language-model evaluation, with influential works including FlexMatch and domain-generalization research.

Jindong Wang’s publication record spans machine learning, computer vision, natural language processing, federated learning, domain generalization, and foundation-model research. Selected highly cited works include a survey of large language model evaluation, a survey of domain generalization, FlexMatch for semi-supervised learning, deep subdomain adaptation networks, and FedHealth for federated transfer learning.[1] [2] [3] [4] [5]

1. Chang, Y., Wang, X., Wang, J., et al. (2024). A survey on evaluation of large language models. ACM Transactions on Intelligent Systems and Technology.
2. Wang, J., Lan, C., Liu, C., Ouyang, Y., Qin, T., Lu, W., Chen, Y., Zeng, W., & Yu, P. S. (2022). Generalizing to unseen domains: A survey on domain generalization. IEEE Transactions on Knowledge and Data Engineering, 35(8), 8052–8072.
3. Zhang, B., Wang, Y., Hou, W., Wu, H., Wang, J., Okumura, M., & Shinozaki, T. (2021). FlexMatch: Boosting semi-supervised learning with curriculum pseudo labeling. NeurIPS 2021.
4. Zhu, Y., Zhuang, F., Wang, J., Ke, G., Chen, J., Bian, J., Xiong, H., & He, Q. (2020). Deep subdomain adaptation network for image classification. IEEE Transactions on Neural Networks and Learning Systems, 32(4), 1713–1722.
5. Chen, Y., Qin, X., Wang, J., Yu, C., & Gao, W. (2020). FedHealth: A federated transfer learning framework for wearable healthcare. IEEE Intelligent Systems.

Jindong Wang is also the author of the monograph Introduction to Transfer Learning: Algorithms and Practice, co-authored with Yiqiang Chen and published by Springer Nature in 2023, and contributed the chapter “Activity Recognition” to the Machine Learning for Data Science Handbook. His teaching and tutorial activities further connect the publication program with graduate education and professional dissemination.

Research Impact

The reported scholarly metrics indicate substantial bibliometric visibility, with 199 documents, 28,805 citations, and an h-index of 61 and The author has published 135 Scopus-indexed documents, receiving 15,924 citations and achieving an h-index of 42. His selected publications have appeared in venues including NeurIPS and major IEEE and ACM journals, while his work has also been represented through open-source projects and research tutorials. Citation counts should be interpreted as quantitative indicators of scholarly attention rather than as a complete measure of research quality or societal value. [1]

Reported open-source adoption includes more than 13,000 stars for the Transfer Learning repository, approximately 2,500 for PromptBench, and more than 1,400 for USB, alongside smaller but specialized communities around RobustLearn and PersonalizedFL. These resources support reproducible experimentation and provide researchers with implementations and evaluation infrastructure.

Media and professional dissemination have included coverage or discussion involving Forbes, MIT Technology Review, Microsoft Research, PyTorch-related channels, TechXplore, LlamaIndex, and other technology publications. Invited and keynote talks have addressed foundation models, LLM evaluation, trustworthy AI, multimodal models, agentic systems, and AI safety across universities, conferences, and research forums.

Award Suitability

The documented profile presents several factors relevant to an innovative research recognition: a sustained publication record, significant citation activity, contributions spanning multiple machine-learning areas, open-source research infrastructure, academic and professional service, teaching, and interdisciplinary engagement. The record also lists competitive research awards and grants from organizations including Google, NVIDIA, Amazon, Microsoft, AMD, Cohere Labs, Modal, and William & Mary.[1]

• Research distinction: substantial publication and citation activity across established machine-learning research areas.
• Innovation: development of methods, benchmarks, surveys, and open-source tools addressing emerging machine-learning problems.
• Community contribution: open-source repositories, tutorials, academic service, and research dissemination.
• Contemporary relevance: active research on foundation models, LLM evaluation, multimodal AI, personalization, robustness, and responsible AI.

Awards and grants reported in the profile include the Gemini Academic Research Award, Google TPU Machine Learning Research Award, AAAI Outstanding SPC Award, NVIDIA Academic Grant Program Award, Amazon Research Award, Google DeepMind Research Award, Microsoft Accelerate Foundation Model Research Grant, AMD University Program AI & HPC Award, Cohere Labs Research Grant, and several best- or outstanding-paper distinctions.

Conclusion

Jindong Wang’s academic profile combines research in machine learning foundations with contemporary work on large language models, foundation models, federated learning, multimodal systems, and responsible AI. The combination of publications, citations, open-source projects, academic service, teaching, awards, grants, and invited research activities provides a broad basis for evaluating his suitability for an innovative research recognition.[1]

External Links

• Scopus Author Profile
• ORCID Profile
• Google Scholar Profile
• Top Teachers Awards

References

1. Google Scholar (n.d.). Google Scholar author Citations details: Jindong Wang,. https://scholar.google.com/citations?user=hBZ_tKsAAAAJ&hl=en&oi=sra
2. Wang, J., Lan, C., Liu, C., Ouyang, Y., Qin, T., Lu, W., Chen, Y., Zeng, W., & Yu, P. S. (2022). Generalizing to unseen domains: A survey on domain generalization. IEEE Transactions on Knowledge and Data Engineering, 35(8), 8052–8072. https://doi.org/10.1109/TKDE.2022.3178128
3. Zhang, B., Wang, Y., Hou, W., Wu, H., Wang, J., Okumura, M., & Shinozaki, T. (2021). FlexMatch: Boosting semi-supervised learning with curriculum pseudo labeling. Advances in Neural Information Processing Systems (NeurIPS 2021). https://proceedings.neurips.cc/paper_files/paper/2021/hash/995693c15f439e3d189b06e89d145dd5-Abstract.html
4. Zhu, Y., Zhuang, F., Wang, J., Ke, G., Chen, J., Bian, J., Xiong, H., & He, Q. (2020). Deep subdomain adaptation network for image classification. IEEE Transactions on Neural Networks and Learning Systems, 32(4), 1713–1722. https://doi.org/10.1109/TNNLS.2020.2988928
5. Chen, Y., Qin, X., Wang, J., Yu, C., & Gao, W. (2020). FedHealth: A federated transfer learning framework for wearable healthcare. IEEE Intelligent Systems. https://doi.org/10.1109/MIS.2020.2988604