Research

17 experimental language-model architectures where hidden dim = vocab size — no embedding, no output projection; every intermediate state is a distribution over vocabulary.

BCS Thesis: Reinforced Linear Genetic Programming

A deep learning system combining transformers and CNNs to classify diseases from patient-described symptoms, achieving 90% recall through semantic embeddings and CNN feature extraction.

A CNN research project exploring optimal image classification architectures for small datasets, demonstrating that data quantity remains the fundamental bottleneck for deep learning performance.