Lensa ML
Guided routes

Follow a path, start to finish

Curated sequences that take you from the basics to the frontier — each step builds on the one before it.

Guided journeys
Path 01 · 7 steps

ML from Zero

Build up from a single neuron to training your first network

  1. 1Neural Networks
  2. 2Forward Propagation
  3. 3Activation Functions
  4. 4Derivatives
  5. 5The Chain Rule
  6. 6Loss Functions
  7. 7Backpropagation
Path 02 · 7 steps

Deep Learning Core

Master the training loop — optimizers, regularization, and normalization

  1. 1Gradient Descent
  2. 2Learning Rate Scheduling
  3. 3Optimizers
  4. 4Batch Normalization
  5. 5Weight Initialization
  6. 6Vanishing & Exploding Gradients
  7. 7Regularization
Path 03 · 7 steps

How LLMs Work

From attention to decoder-only transformers — the architecture behind GPT

  1. 1Softmax, Logits & Temperature
  2. 2Embeddings
  3. 3How Self-Attention Works
  4. 4Multi-Head Attention
  5. 5Positional Encoding
  6. 6The Transformer Architecture
  7. 7Decoder-Only Transformers
Path 04 · 6 steps

Math Foundations

The calculus and linear algebra you need for deep learning

  1. 1Derivatives
  2. 2The Chain Rule
  3. 3Matrix Calculus & Jacobians
  4. 4Linear Algebra
  5. 5Tensors, Broadcasting & Reshaping
  6. 6Eigenvalues & Eigenvectors
Path 05 · 6 steps

Probability & Stats

Distributions, Bayes, and information theory for ML

  1. 1Probability & Distributions
  2. 2Bayes' Theorem
  3. 3Statistics Essentials
  4. 4MLE & MAP Estimation
  5. 5Entropy & Cross-Entropy
  6. 6KL Divergence
Browse by topic

Every topic, as a path

Shorter routes through one area — pick a subject and work straight through it.

Ready to dive in? Every path is free.