I am a fourth year PhD student in Electrical Engineering at the Universoty of Southern California, working with Prof. Salman Avestimehr. My research focuses on the intersection of federated learning and graph machine learning. Recently, I have a keen interest in generative flow networks and conformal prediciton. I have been selected as a 2025 North America Finalist at the Qualcomm Innovation Fellowship.

Research Topics

Federated Learning

GFlowNets

Graph Neural Networks

Uncertainty Estimation

Probabilistic Modeling

AI For Science

Conformal Prediction

LLM Reasoning

My Contributions +

As AI systems are increasingly deployed in complex tasks and real-world applications, I believe it's crucial to understand and improve their reasoning and reliability. My work addresses challenges like quantifying uncertainty, performing efficient search and sampling, and ensuring reliable reasoning, all of which are fundamental for safe and effective AI deployment.

One major focus has been the development of Generative Flow Networks (GFlowNets), a framework that connects reinforcement learning and probabilistic modeling. I've worked on establishing the theoretical foundations of GFlowNets, demonstrating their effectiveness in Bayesian inference of discrete and continuous structures, and showing their ability to handle multimodal distributions better than traditional variational methods. I also created and released torchgfn, an open-source software framework, to make GFlowNets more accessible for both discrete and general applications.

Beyond GFlowNets, I've also contributed to uncertainty estimation through a method called DEUP (Direct Epistemic Uncertainty Prediction). DEUP addresses model misspecification in interactive learning settings by using a secondary predictor trained on the main model's errors. This results in more reliable uncertainty estimates for decision-making in downstream tasks.

Currently, I'm focused on the theory and applications of GFlowNets to tackle more complex problems, using them to enhance Large Language Model reasoning, and developing even better uncertainty estimation methods. My overall goal is to improve the reliability, efficiency, and reasoning capabilities of AI systems, particularly in scenarios that demand robust decision-making under uncertainty. For this goal, I am collaborating with Salem Lahlou and Saurav Prakash.

Recent News

• May 2025 - FALCON, an end-to-end ML framework for analog circuit design (including topology selection, layout-aware parameter selection, and performance prediction) is out!
• February 2025 - Became a finalist at the 2025 Qualcomm Innovation Fellowship.
• February 2025 - Awarded with Travel Grant for the SIAM-SDM'25 conference, a top-tier conference in ML & data-mining..
• January 2025 - Became a semi-finalist at the 2025 Qualcomm Innovation Fellowship .
• December 2024 - FedGrAINS, first GFlowNet paper to improve subgraph federated learning has been accepted to the SIAM-SDM'25 conference. Preprint is available in this link.

Ultimately, I am driven by a deep interest in understanding and defining intelligence, whether it's in animals or artificial systems. While I'm excited about the potential of AI, I'm also mindful of its limitations and potential pitfalls – I'm definitely not a technosolutionist.

Before my current role, Before joining USC, I received my MSc and BSc degrees from the Electrical Engineering at Bilkent University as a Senior Researcher in 2018 and 2020 respectively.