About me
I am currently a Principal Scientist at Amazon, where I spearhead the development of next-generation foundational models for Amazon Stores businesses. Prior to this role, I held the position of Principal Researcher at Tencent AI Lab, specializing in the advancement of large language models. Before that, I have been working at Microsoft Research in Redmond, WA, where I delved into deep learning, natural language processing, and reinforcement learning. I earned my PhD from University of California, Los Angeles (UCLA), in June 2014, focusing on distributed learning and control.
Contact information
Email: chenjianshu at gmail dot com
Research interests
My research interests lie at the intersection of machine learning, natural language processing, and large language models. I focus on understanding and optimizing the synergy between knowledge and reasoning to develop next-generation large language model architectures and effective learning paradigms, with the objective of achieving strong compositional generalization, reasoning and planning capabilities. I am passionate about tackling large-scale AI research projects, collaborating with interdisciplinary teams to address complex challenges, and driving robust and effective innovations in AI. Additionally, I maintain an active interest in reinforcement learning and optimization.
For more details, see my publications (also google scholar)
Selected publications
- R. Yang, X. Pan, F. Luo, S. Qiu, H. Zhong, D. Yu, Jianshu Chen, “Rewards-in-Context: Multi-objective Alignment of Foundation Models with Dynamic Preference Adjustment”, Proc. International Conference on Machine Learning (ICML), July 2024.
- X. Zhao, H. Zhang, X. Pan, W. Yao, D. Yu, T. Wu, Jianshu Chen, “Fact-and-Reflection (FaR) Improves Confidence Calibration of Large Language Models”, arXiv preprint [arXiv:2402.17124], February, 2024.
- J. Chen, X. Pan, K. Song, D. Yu, D. Yu, Jianshu Chen, “Skills-in-Context Prompting: Unlocking Compositionality in Large Language Models”, arXiv preprint [arXiv:2308.00304], August 2023.
- Jianshu Chen, “Learning Language Representations with Logical Inductive Bias”, Proc. International Conference on Learning Representations (ICLR), 2023.
- X. Pan, W. Yao, H. Zhang, D. Yu, D. Yu, Jianshu Chen, “Knowledge-in-Context: Towards Knowledgeable Semi-Parametric Language Models”, Proc. International Conference on Learning Representations (ICLR), 2023 (Spotlight).