Curriculum Vitae

University of Maryland, College Park

• GPA: 3.93/4.00
• Advisor: Prof. Sennur Ulukus
• Focus: Information Theory, Interpretability of Large Language Models
• Expected graduation: Aug. 2030

Bilkent University

• GPA: 3.92/4.00

National University of Singapore

• Relevant coursework: Quantum Communication and Cryptography, Quantum Mechanics I, Probability

Graduate Research Assistant

• Develop information-theoretic and dynamical-systems tools for compressing and interpreting transformer architectures, spanning task-aware rate allocation and mechanistic analysis of attention and feed-forward blocks; one conference paper accepted, one journal submission under review, and two manuscripts in preparation within the first year of the program.
• Collaborate across four institutions (UMD, UW–Madison, Oxford, Ohio State) on the mathematical foundations of deep learning, translating abstract dynamical characterizations into empirically testable claims about model behavior.

Undergraduate Researcher

• Built the embedded hardware demonstrator that turned the group’s theoretical semantic signal processing framework into a working multi-sensor system, establishing the experimental basis for a journal publication.
• Profiled semantic extraction under strict power and memory budgets on Raspberry Pi, quantifying the compute floor required for real-time goal-oriented inference at the edge.
• Implemented a method-of-moments solver from first principles to model plane-wave scattering by a perfectly conducting strip, validating numerical results against analytical limits.

Summer Intern

• Designed a suspended ring resonator model in MATLAB and CST Studio that estimates dielectric constant and loss tangent, delivering a low-cost alternative to commercial material characterization for antenna development.
• Cross-validated MATLAB against full-wave CST simulations to bound estimation error, then generalized the method into a lookup table and polynomial fits so the tool transfers across materials without re-derivation.

R&D Intern

• Prototyped a two-node semantic communication network on Raspberry Pi devices that performs visual semantic extraction entirely on-device, collapsing the transmission requirement from a 10 Mbps raw video stream to roughly 10 bps of task-relevant semantic tokens, a ~10⁵× reduction in bandwidth achieved by transmitting meaning rather than pixels.
• Engineered the TCP-based transport layer and a stream fusion algorithm that merged asynchronous multi-sensor inputs, improving integration accuracy by 67.2% over unfused processing while operating within the power and memory envelope of embedded hardware.

Summer Intern

• Built a multithreaded, socket-based file transfer system in C++ supporting arbitrary file types via binary-mode I/O, applying factory and observer patterns to keep socket creation and worker notification decoupled and extensible.

The Rate–Distortion–Deception Tradeoff

Practical Hardware Demonstration of a Multi-Sensor Goal-Oriented Semantic Signal Processing and Communications Network

On the Diverse Dynamical Behaviors Arising in Deep Linear Transformers

Task-aware Compression with Distribution Matching

The Role of Feed-Forward Layers in Transformer Dynamics

Academic Excellence Award

Awarded to top-ranked students for sustained academic distinction across the degree program.

Research Excellence Award

Recognized for research contributions culminating in a first-authored journal publication as an undergraduate.

Best Project Award

Selected department-wide for excellence and originality in project design and execution.