Lab
Where my research lives
Every page here starts with a question I couldn't stop thinking about and ends with numbers someone else can check. The method is written out, the figures are drawn from the results, and the files sit underneath.
01
Research
Original research, ranked by an expert panel, with the code.
02
Writeups
A question, the method, and what the numbers said. Code included.
- Sep 2026When you squeeze a webpage, do the boxes move in a pattern?WriteupI wanted to know if a webpage has any physics. When you make the window narrower, do the boxes move in a pattern a model can learn, or is every page a mess? I measured 3,056 boxes on 25 pages. A one-line CSS rule beat a simple linear model. A tree model beat the rule. Median error went from 26.7 pixels to 7.3.
- Sep 2026A Pokémon card agent that plays itselfKaggle, The Pokémon Company PTCG AI BattleI built an agent that plays Pokémon TCG in the cabt simulator. I mined daily match logs for winning decks, then made the agent play itself. Best ladder score 708. Rank 2057 of 6807.
- Aug 2026Guess the rock along a drill pathKaggle, ROGII Wellbore Geology PredictionROGII wellbore geology. I guessed TVT, the depth marker along a horizontal well, with LightGBM, CatBoost, and Ridge on 5 folds grouped by well. Best of my accepted rows 13.894. Rank 4696 of 6125.
- June 2026Teach the steps, not the last numberKaggle, NVIDIA Nemotron Model Reasoning ChallengeNVIDIA Nemotron reasoning. Each puzzle hides a tiny program. I fitted that program, wrote gold step-by-step traces from it, then LoRA-tuned Nemotron-3-Nano. Public 0.704, private 0.720. Rank 2588 of 4182.
- May 2026A math model that runs on the same machineKaggle, AI Mathematical Olympiad Progress Prize 3AIMO Progress Prize 3. I ran gpt-oss on a local vLLM server with Harmony prompts, and I had a Qwen2.5-Math pipeline next to it. Private 38.5. Rank 2345 of 4138.
- May 2026Will they pit on the next lap?Kaggle, Predicting F1 Pit StopsPlayground Series S6E5. I mixed CatBoost, XGBoost, and LightGBM over 10 folds to guess whether a driver pits next lap. Private AUC 0.950. Rank 1090 of 3022.
03
Papers
Shorter pieces: positions, critiques, and ideas I'm testing.
- Aug 2026The floor is worth raisingOn Taste Labs, Requests for research, 16 Aug 2026I read Taste Labs' research request the week it went up. This is me answering it in public so I can't pretend I didn't.
- Aug 2026Taste is the thing we actually hireNoteA chair that holds you and a chair you want in the room are different jobs. We keep scoring the first one.
- Aug 2026If it only knows the data, put taste in the dataNoteA model is what you feed it. We keep feeding it the average of the web and then asking why it has no pulse.