Explore
Technical explainers on applied AI/ML
Longform writing on model mechanics, evaluation frameworks, training dynamics, and system architecture, with the assumptions made visible.
A Better Way to Write Project Notes
A project page earns its space when a reader can follow the problem, the engineering choices, and the evidence. Here…
Reproducing a Paper Starts With Its Assumptions
A method section is more than a recipe. Reading its assumptions closely helps distinguish a faithful reproduction from a result…
A Model Score Is Not a Decision
A strong metric can still hide the failure that matters in practice. Start evaluation by asking what decision the model…