reading

Reading

Two ground-up courses. Every sentence built to be clear.

Scaling Laws, Made Simple

A 10-part ground-up course on the 2020 'Scaling Laws for Neural Language Models' paper: powers of ten, power laws, log-log plots, loss, the three levers, and how to spend a compute budget.

10 parts

Start reading →

Agent Memory, Made Clear

A 13-part course on memory in agentic and LLM systems: the four memory types, storage, retrieval, writing and forgetting, modern systems (Letta, Mem0, Zep), optimization, evaluation, security - and a product blueprint.

13 parts

Start reading →

The one-page hand-drawn map of agent memory →