John Hongyu Meng
Ph.D., Principal Investigator
Email: menghongyu@@gdiist.cn
Personal Profile
Hongyu Meng is the Principal Investigator and Leader of the Predictive Coding and Cognitive Systems lab. He received his B.S. in Applied Mathematics from Peking University in 2013 and his Ph.D. in Applied Mathematics from Northwestern University in 2019. He subsequently conducted postdoctoral research at the Center for Neural Science at New York University.
Dr. Meng’s research focuses on the neural mechanisms of predictive coding and brain-inspired intelligence, with particular interests in the neural computation of predictive coding, neural dynamics, neuronal cell types and their functional roles, and oscillatory mechanisms in spiking neural networks. His research has been published in international journals including Nature Communications, eLife, Journal of Neuroscience, PLOS Computational Biology, and Scientific Reports.
Laboratory of Predictive Coding and Cognitive Systems
Our lab focuses on the mechanisms and technological applications of predictive coding in neuroscience and brain-inspired intelligence. More specifically, the ability to detect “surprise” is fundamental to perception, learning, and adaptive behavior, and represents a central question in predictive coding theory. In neuroscience, our lab investigates how the brain forms and communicates internal representations, and how these representations influence perception and prediction across different brain regions. In brain-inspired intelligence, we explore how predictive coding mechanisms can be integrated with spiking neural networks to improve the computational efficiency of large-scale neural models, while advancing their deployment on neuromorphic hardware.
Our current research focuses on two major directions:
1. Neural mechanisms of predictive coding: Building on our existing theoretical frameworks, publicly available neuroscience datasets, and collaborations with experimental neuroscience labs, we use data-driven approaches to test the core hypotheses of predictive coding. We aim to develop computational models and theoretical frameworks that are consistent with experimental observations and to uncover how predictive coding is implemented across brain regions and neural circuits.
2. Brain-inspired algorithms based on predictive coding: The human brain relies on massively parallel spiking computation to respond to external stimuli on extremely short timescales (<50 ms) and to rapidly adapt to changing environmental and task contexts. Guided by predictive coding mechanisms identified through neuroscience research, our group aims to develop a new generation of fast, low-power, and energy-efficient brain-inspired algorithms, with the goal of implementing them in spiking neural networks and deploying them on brain-inspired processing units (BPUs).
Our group is continuously recruiting postdoctoral researchers (Postdoc), research assistants (RA), and interns in the following areas: (1) computational neuroscience, (2) AI for Science (AI4Science), and (3) neuroscience data visualization and modeling. Interested candidates are encouraged to contact us by email.
Selected Publications:
(1) Meng JH, Wang X-J. Global error signal guides local optimization in mismatch calculation. Nature Communications. 2026, Mar 12;17:3868.
(2) Meng JH, Ross JM, Hamm JP, Wang X-J. Duet model unifies diverse neuroscience experimental findings on predictive coding. BioRxiv. 2025.
(3) Aizenbud I, Audette N, Auksztulewicz R,... Meng JH*,..., Xiong, Y. S. Neural mechanisms of predictive processing: a collaborative community experiment through the OpenScope program. (Authors ordered by last name) arXiv. 2025.
(4) Meng JH*, Kang Y*, Lai A, Feyerabend M, Inoue W, Martinez-Trujillo J, Rudy B, Wang X-J. In Search of Transcriptomic Correlates of Neuronal Firing-Rate Adaptation across Subtypes, Regions, and Species: A Patch-seq Analysis. BioRxiv. 2024.
(5)Meng JH, Schuman B, Rudy B, Wang X-J. Mechanisms of dominant electrophysiological features of four subtypes of layer 1 interneurons. Journal of Neuroscience. 2023 May 3;43(18):3202-18.
(6) Meng JH, Riecke H. Structural spine plasticity: Learning and forgetting of odor-specific subnetworks in the olfactory bulb. PLoS Computational Biology. 2022 Oct 24;18(10):e1010338.
(7) Meng JH, Riecke H. Synchronization by uncorrelated noise: interacting rhythms in interconnected oscillator networks. Scientific Reports 2018;8.1:6949.
(8) Dellal S, Zurita H, Valero M, Abad-Perez P, Kruglikov I, Meng JH, Prönneke A, Hanson JL, Mir E, On-garo M, Wang X-J, Buzsáki G, Machold R, Rudy B. Inhibitory and disinhibitory VIP IN-mediated circuits in neocortex. BioRxiv. 2025.
(9) Machold R, Dellal S, Valero M, Zurita H, Kruglikov I, Meng JH, Hanson JL, Hashikawa Y, Schuman B, Buzsáki G, Rudy B. Id2 GABAergic interneurons comprise a neglected fourth major group of cortical inhibitory cells. eLife. 2023 Sep 4;12:e85893.
