Thoughts on mathematics, cryptography, decentralized systems, machine learning, and whatever I'm exploring at the moment.
Two AI teams play mus — the Spanish card game whose rulebook allows partners to speak only through facial gestures. An órdago accepted on a tiebreak rule, a quarter of the signal channel lost to timing, and the harness catching an error I had already written into a draft.
My master thesis exploring advanced reinforcement learning techniques and their applications. Featuring implementations of Soft Actor-Critic (SAC) algorithm and comprehensive analysis of modern RL methods in continuous control tasks.
Implementation of a Kolmogorov-Arnold neural network using Taylor polynomial expansion in PyTorch. Exploring novel approaches to function approximation and deep learning architectures.
Introducing a bayesian approach to determine the label of a data example based on the labels provided by multiple noisy annotators. In progress