Here, I test how the mathematical theory of generalization bounds works in practice.
Here, I implement a decoder-only language model using Pytorch, and I build a trainer to train the model on a shakespeare dataset.
Here, I implement PPO and show it working for breakout and cartpole.
Here, I implement RLHF using my own transformer implementation and a simple policy gradient algorithm.
This is an opinion I wrote for the Journal of Undergraduate Science and Technology in Fall 2023. It details computer-assisted proofs from early computers to new technologies like language models.
Here, I use a CNN to classify cats and dogs and integrate it with a Flask web app.