Cheng-Yeh Chen

Tokyo, Japan, 2025

I am a Ph.D. student at Georgia Tech ECE, advised by Prof. Raghupathy Sivakumar in the Georgia Tech GNAN lab. I study how brains and AI models experience the world in surprisingly similar ways — and how each can unlock new capabilities in the other. My research sits at the intersection of neuroscience and deep learning, organized around three themes:

Brain-machine alignment (🧠🔄🤖) — Mapping the computational parallels between transformers and the brain.
Transformer interpretability (🧠➡️🤖) — Using the brain’s functional topography as a lens to understand and steer what vision and language models learn.
Neural decoding (🧠⬅️🤖) — Reconstructing perception from brain signals (fMRI, EEG, MEG) via brain-machine alignment and transformer interpretability.

I received my B.S. and M.S. degrees in Electrical Engineering and Communication Engineering from National Taiwan University. I was fortunate to work with Prof. Hung-Yun Hsieh in the TONIC lab on 5G security, wireless virtual reality, and immersive video streaming.

selected publications

ICLRTop 7% Reviews

The Mind’s Transformer: Computational Neuroanatomy of LLM-Brain Alignment

Cheng-Yeh Chen, and Raghupathy Sivakumar

In International Conference on Learning Representations (ICLR) 2026

Website OpenReview Code Bib

@inproceedings{mindtransformer,
  author = {Chen, Cheng-Yeh and Sivakumar, Raghupathy},
  booktitle = {International Conference on Learning Representations (ICLR)},
  title = {The Mind's Transformer: Computational Neuroanatomy of LLM-Brain Alignment},
  year = {2026},
}

TWC

Cross-Frame Resource Allocation with Context-Aware QoE Estimation for 360° Video Streaming in Wireless Virtual Reality

Cheng-Yeh Chen, and Hung-Yun Hsieh

IEEE Transactions on Wireless Communications 2023

Bib HTML Supp

@article{twc,
  author = {Chen, Cheng-Yeh and Hsieh, Hung-Yun},
  journal = {IEEE Transactions on Wireless Communications},
  title = {Cross-Frame Resource Allocation with Context-Aware QoE Estimation for 360° Video Streaming in Wireless Virtual Reality},
  year = {2023},
  volume = {22},
  number = {11},
  pages = {7887-7901},
  doi = {10.1109/TWC.2023.3257059},
}

news

Jan 26, 2026	Our work, “The Mind’s Transformer: Computational Neuroanatomy of LLM-Brain Alignment,” is accepted to ICLR 2026!
Aug 1, 2025	We are ranked 5th globally (out of 53 teams) in the NeurIPS 2025 LibriBrain — MEG Speech Detection challenge (F1: 0.9071; 1st place: 0.9166)!
Oct 21, 2024	Our work, “Data-Centric Resource Allocation for Machine-Type Communications with Lossy Links Based on Compressive Sensing,” is accepted to IEEE Transactions on Vehicular Technology!
Oct 2, 2024	Our work, “Uplink performance analysis and optimization for dense clustered wireless networks with nearest-BS association,” is accepted to IEEE Transactions on Communications!
Jan 17, 2024	Our work, “Towards Optimal Multiview Transcoding for Edge-Assisted Wireless Volumetric Streaming,” is accepted to IEEE International Conference on Communications 2024!