
@omar
Joined November 2026
Roadmaps
4

Machine Learning from First Principles
12 weeks · 1 follower
Notes
32Diffusion: noise in, image out
Deep Learning from Scratch: backprop to diffusion · Nov 2, 2026Superposition and sparse autoencoders
Large Language Models, End to End · Nov 2, 2026Reading a confusion matrix and choosing a threshold
Machine Learning from First Principles · Nov 2, 2026Prompt injection is a confused-deputy problem
Building AI Agents: tools, memory and evals · Nov 2, 2026- Large Language Models, End to End · Nov 2, 2026
Data leakage: five ways to fool yourself
Machine Learning from First Principles · Nov 2, 2026- Building AI Agents: tools, memory and evals · Nov 2, 2026
- Deep Learning from Scratch: backprop to diffusion · Nov 2, 2026
- Building AI Agents: tools, memory and evals · Nov 2, 2026
- Machine Learning from First Principles · Nov 2, 2026
- Deep Learning from Scratch: backprop to diffusion · Nov 2, 2026
- Large Language Models, End to End · Nov 2, 2026
- Building AI Agents: tools, memory and evals · Nov 2, 2026
The KV cache and why inference is memory-bound
Large Language Models, End to End · Nov 2, 2026How gradient boosting fits residuals
Machine Learning from First Principles · Nov 2, 2026- Deep Learning from Scratch: backprop to diffusion · Nov 2, 2026
When retrieval beats a longer context
Building AI Agents: tools, memory and evals · Nov 2, 2026How a decision tree chooses a split
Machine Learning from First Principles · Nov 2, 2026Why scaling laws bend (Chinchilla)
Large Language Models, End to End · Nov 2, 2026Residual connections and gradient flow
Deep Learning from Scratch: backprop to diffusion · Nov 2, 2026- Large Language Models, End to End · Nov 2, 2026
- Deep Learning from Scratch: backprop to diffusion · Nov 2, 2026
The bias-variance decomposition with real numbers
Machine Learning from First Principles · Nov 2, 2026Designing tools a model can use
Building AI Agents: tools, memory and evals · Nov 2, 2026Why cross-entropy and not squared error
Machine Learning from First Principles · Nov 2, 2026- Building AI Agents: tools, memory and evals · Nov 2, 2026
Why initialization matters: the variance math
Deep Learning from Scratch: backprop to diffusion · Nov 2, 2026Attention, one matrix at a time
Large Language Models, End to End · Nov 2, 2026Backprop is the chain rule with bookkeeping
Deep Learning from Scratch: backprop to diffusion · Nov 2, 2026Gradient descent, derived and drawn
Machine Learning from First Principles · Nov 2, 2026- Large Language Models, End to End · Nov 2, 2026
- Building AI Agents: tools, memory and evals · Nov 2, 2026

