OpenAI patched a verbal tic out of their model. The patch ships in Codex via a developer prompt, not in Copilot, so in Copilot the goblins still appear. I ran the experiment I was curious about: does goblin mode actually change coding performance, or is it just vibes?
Read more →
The heaviest users of my command-line tools turned out to be AI agents. Here's what I changed once I noticed.
Read more →
A nine-hour experiment on why reasoning models are so stubborn about correcting their own chain of thought, and a latency finding I didn't expect.
Read more →
My senior capstone was a flood-warning system for Vietnam. Most of the actual work was in everything around the code — stakeholders, data ownership, and ethics.
Read more →
As part of the Hacking For Defense class at UVA, my team provides reccomendations to realign the US Cyber Budget.
Read more →
A summer internship project. Most explainability methods try to reverse-engineer a model after the fact. I tried training one to just tell you why.
Read more →
A college information-retrieval project, a dataset of pro-ISIS accounts, and what happened when we tried to break our own ranking algorithm with Kanye West.
Read more →