Archive · 2020–2026
Coursework and selected implementations from Berkeley and TJHSST.
Feature-based text classification, tokenization, transformer language models, training, prompting, and retrieval-augmented question answering.
Work Classifiers; BPE tokenizer and transformer; offline QA system.
Internet measurement, packet forwarding, routing, and end-to-end transport protocols.
Work Traceroute; distance-vector routing; reliable transport.
State-space search, adversarial agents, probabilistic tracking, reinforcement learning, and neural models.
C, RISC-V assembly, memory hierarchy, digital logic, and processor design.
Work Snake in C; RISC-V neural-network routines; RISC-V CPU.
Strategic games, auctions, economic design, and the algorithmic study of incentives.
Archive Lecture material and six problem sets.
ResearchRamp, a multi-agent literature-mapping tool that identifies foundational topics, survey papers, and seminal work, then renders an interactive research graph.
Chattom, a framework for evaluating conversations across models, personas, and different conversation lengths.
Proofs, discrete structures, computability, probability, random variables, and concentration.
Archive Twenty-three course notes.
Data structures, algorithms, testing, and software design in Java.
Work 2048; deques; Gitlet version control; tile-based world generation.
Functional abstraction, recursion, higher-order functions, mutable state, and object-oriented programming in Python.
Work Hog; Cats typing test; Ants vs. SomeBees.
Bayesian methods, dimensionality reduction, classification, neural networks, and MNIST experiments.
BFS, DFS, A*, constraint satisfaction, game agents, genetic algorithms, perceptrons, and neural networks.
Android development through authentication, chat, a custom-drawn game, and a task-management final project.
HTML, CSS, JavaScript, Bootstrap, external APIs, cookies, and authentication through small web applications.