Projects
Neural networks, evolution and simulations, built from scratch.
01. Transformer Language Model
Deep Learning · NLP · Transformers · PyTorch
Designed and trained a decoder-only Transformer language model from random initialization, implementing causal self-attention, autoregressive generation, tokenization and model interpretability tools.
02. Neuroevolutionary Driving Intelligence
Neuroevolution · Robotics · Simulation · TypeScript
Built a physics-based autonomous driving simulation where neural-network controllers evolve through genetic optimization to navigate unfamiliar tracks.
03. Artificial Life & Evolutionary Intelligence
Artificial Life · Multi-Agent Systems · Evolutionary Computing
Developed a digital ecosystem of autonomous neural organisms with inherited genomes, mutation, environmental selection and emergent population dynamics.
04. Neural Network Research Laboratory
Neural Networks · NumPy · Backpropagation · Optimization
Implemented a configurable multilayer neural network with manual backpropagation, gradient verification, live decision-boundary visualization and real-time training diagnostics.
05. Neural Vision Engine (NVE)
Computer Vision · CNNs · PyTorch · Model Interpretability
Developed a locally trained convolutional vision model for handwritten-digit recognition, with interactive feature-map inspection and model diagnostics.