Joel Ye
- Program Year: 3
- Academic Institution: Carnegie Mellon University
- Field of Study: Computational Neuroscience
- Academic Advisor: Leila Wehbe
- Practicum(s):
Lawrence Berkeley National Laboratory (2024) - Degree(s):
M.S. Computer Science, and B.S. Computer Science, Georgia Institute of Technology, 2020
Summary of Research
My research aims to relate computation in the brain and in AI systems, and to develop deep learning systems for bi-directional brain-computer interfaces. Specifically, I aim to design neurostimulation models to equip brain-computer interfaces with more functionally useful and naturalistic sensation.Publications
Neural Data Transformer 2: Multi-context Pretraining for Neural Spiking Activity. Neural Information Processing Systems (NeurIPS), 2023. J. Ye*, J. Collinger, L. Wehbe, R. Gaunt.Neural Latents Benchmark: Evaluating latent variable models of neural population activity. Neural Information Processing Systems (NeurIPS) Benchmarks and Datasets, 2021. F. Pei*, J. Ye*, D. Zoltowski, A. Wu, R. Chowdhury, H. Sohn, J. O'Doherty, K. Shenoy, M. Kaufman, M. Churchland, M. Jazayeri, L. Miller, J. Pillow, M. Park, E. Dyer, C. Pandarinath.
Auxiliary Tasks and Exploration Enable ObjectNav. International Conference on Computer Vision (ICCV) 2021. J. Ye, D. Batra, A. Das, and E. Wijmans.
Auxiliary Tasks Speed Up Learning PointGoal Navigation. Conference on Robot Learning (CoRL), 2020. J. Ye, D. Batra, E. Wijmans, and A. Das.
Representation learning for neural population activity with Neural Data Transformers. Neurons, Behavior, Data analysis, and Theory (NBDT), 2021. J. Ye, C. Pandarinath.
I have given accompanying talks in the associated conferences, along with informal presentations for journal clubs for the above works.
Representation learning for neural population activity with Neural Data Transformers (an earlier version of the paper above) was presented at Neuromatch 3.0.