Explainer

Five methods for planning a working week with a language model, from multi-dimensional prioritization to using the model as a critic.

A system prompt built in layers, from success criteria to role, context and verification, with every step shown as working text.

Habits for checking what a model gives you: classify the claim, compare models, make the model attack its answer, verify every citation.

Raters preferred a 1.3-billion-parameter model over one with 175 billion. Alignment, RLHF, and why no answer comes out neutral.

Push a model past a perfect fit on its training data and test error starts falling again. Double descent, and what it did to learning theory.

Phi-3 Mini runs on a laptop and beats models twice its size on Microsoft's own table. Why model size stopped being the first question.