Yehya Farhat
I am a second year PhD student in the Computer Science department at Rice University, advised by Anastasios Kyrillidis. I am also part of the AI OWLS Ken Kennedy institute research cluster.
My overarching research goal is to develop scalable, rigorous, and efficient methods for training and deploying large-scale AI systems. My work spans several interconnected areas: convex and non-convex optimization, distributed training, model compression, Mixture-of-Experts, and continual learning. I am also keenly interested in leveraging AI agents in real-world settings to solve critical bottlenecks in scientific and industrial domains.
I was a research scholar at Syracuse University, where I worked with Prof. Venkata Gandikota. I previously earned an M.Sc. in Computer Science from Syracuse University, where I completed a thesis under the supervision of Prof. Ferdinando Fioretto. Before that, I received a B.Sc. in Computer Science with a minor in Mathematics from the American University of Beirut (AUB).
In my free time, I enjoy playing chess, staying active by playing football, and hitting the gym. I am also an avid writer passionate about sharing intriguing scientific topics I encounter. To stay updated, you can find and follow me on Medium.
News
| Nov 07, 2025 | 📢 Announcement: I’ll be attending NeurIPS 2025 in San Diego from December 2–8. Lets connect! |
|---|---|
| Nov 07, 2025 | Check out our new preprint TwIST: Rigging the Lottery in Transformers with Independent Subnetwork Training! We propose a novel distributed training algorithm that trains subnetworks in parallel to uncover high-performing sparse models that need no fine-tuning. |
| Sep 19, 2025 | Excited to announce that our paper Learning to Specialize: Joint Gating-Expert Training for Adaptive MoEs in Decentralized Settings has been accepted at NeurIPS2025 |