Set up an Obsidian Zettelkasten that Claude Code, Codex, and Hermes Agent maintain with digest, wiki, qna, and lint skills, after Karpathy's LLM Wiki.

An exploration of adaptive join order selection techniques for optimizing multi-way joins in continuous data streams.

Learn how to simulate heat transfer using CUDA, compile and run your code on a Google Cloud GPU VM, and visualize the results remotely using X11 forwarding.

Create a Google Cloud VM with an NVIDIA T4 GPU, connect from VS Code over SSH, compile a first CUDA program, and keep GPU costs under control.

PIXIU, an open-source financial LLM: the FIT instruction-tuning dataset, the FLARE benchmark, and how the fine-tuned model performs on finance NLP tasks.

A geometric picture of Jensen's inequality for convex functions, with a Python plot and what it implies for returns, risk, and option pricing.

Why classifier precision, not accuracy, drives the Sharpe ratio of an ML trading strategy, derived for IID bets with symmetric payouts.

Simulate stock price paths with the Euler-Maruyama scheme in Python and price Asian and lookback options by Monte Carlo under the risk-neutral measure.

A tour of QuantConnect's backtesting, paper trading, and free data, then a portfolio algorithm built from the five Algorithm Framework modules.

How devdocs.io puts NumPy, pandas, and scikit-learn documentation in one searchable tab that also works offline, and why it speeds up quant work.