Just a heads up, the website is still under construction and some parts may be incomplete
About Me
Hey, I’m Seyfal! I’m an undergraduate student at the University of Illinois at Chicago and a research assistant at the Electronic Visualization Laboratory (EVL). At EVL, I work under the supervision of Professors Mike Papka and Lan Zhiling on LASSI, an automated self-correcting pipeline for code generation. I also specialize in optimizing and adapting deep learning workloads for EVL’s Intel GPU cluster, collaborating with researchers to implement and optimize their experiments for Intel’s architecture.
Additionally, I’m conducting research on Spectral Anomaly Detection in EELS Spectral Images using Three-Dimensional Convolutional Variational Autoencoders, under the guidance of Professors Robert F. Klie and James P. Buban.
My research interests include: (i) Communication-efficient distributed training algorithms that minimize cross-node dependencies while preserving model quality, (ii) decentralized architectures that leverage autonomous agents to enable scalable training not limited by centralized compute, and (iii) data-centric optimizations that enhance scaling efficiency through unsupervised preprocessing, representation learning, and continuous knowledge monitoring.
Copy [email protected] View Research Summary →
News
- Our Paper “Robust Spectral Anomaly Detection in EELS Spectral Images via Three Dimensional Convolutional Variational Autoencoders” is on Arxiv and under review at Nature Communications.
- Received High Distinction at UIC Honors College Research Symposium for presentation on “Spectral Anomaly Detection in EELS Spectral Images via Three Dimensional Convolutional Variational Autoencoders” See Poster / See Announcement