← Engineermaxxing
Feynman-Level Interactive Lessons

Gleams

Build intuition from absolute zero. Every concept from first principles, with interactive simulations, step-by-step math, and quizzes. No prerequisites beyond curiosity.

149
Lessons
1256+
Chapters
826+
Simulations
No matches found.
AI Architectures31
LESSON 01

microGPT

From absolute zero to understanding every line of Karpathy's 243-line GPT.

11 chapters15+ sims
LESSON 02

Transformer

Self-attention, multi-head, KV cache, MoE — the architecture behind everything.

11 chapters
NEW

Vision Transformer (ViT)

Split an image into patches, treat them as tokens, run a Transformer. How ViT unified vision and language architectures.

10 chapters10+ sims
NEW

Universal Architecture

Why one design rules everything — retrofitting patterns, cross-attention, conditioning zoo, composition.

10 chapters12+ sims
LESSON 03

SSM / Mamba

Linear recurrence, selective scan — the O(n) alternative to attention.

10 chapters
LESSON 04

Diffusion Models

Add noise, learn to reverse it. The dominant generative paradigm.

10 chapters
LESSON 05

Flow Matching

Straight paths from noise to data — simpler, faster than diffusion.

9 chapters
NEW

Magenta RealTime 2

A 2.4B music model you can jam with in ~200 ms on a laptop — frame-level codec LM, attention sinks, pianoroll & style control.

11 chapters10 sims
NEW

Diffusion Transformer (DiT)

Replace the U-Net with a Transformer. adaLN-Zero conditioning, scaling laws, and how DiT powers SD3, FLUX, and Sora.

10 chapters10+ sims
LESSON 06

VAE / VQ-VAE

The secret plumbing — variational inference, codebooks, tokenization.

10 chapters
LESSON 07

GAN

Generator vs discriminator — the adversarial game.

9 chapters
LESSON 08

Contrastive / CLIP

The glue connecting images and text in a shared space.

10 chapters
LESSON 09

VLM

Teaching language models to see — vision encoder + LLM fusion.

10 chapters
NEW

Pixel Buffers → VLM Tokens

The data layer under every vision model: trace one image from raw bytes through layout, resize, normalization, and patchify all the way to the tokens a VLM reads — and the silent bugs at each step.

10 chapters9 sims
LESSON 10

VLA

Foundation models that physically act in the world.

10 chapters
LESSON 11

World Models

Learning to imagine before acting — prediction as intelligence.

9 chapters
LESSON 12

NeRF / 3DGS

Reconstructing 3D worlds from 2D photographs.

10 chapters
LESSON 13

Reward / Alignment

RLHF, DPO, Constitutional AI — making AI do what we want.

10 chapters
LESSON 14

Agent Evaluation

From vibe checks to rigorous testing. Graders, pass@K/pass^K, τ-bench, Terminal-Bench, the Swiss cheese model.

12 chapters12+ sims
NEW

Mixture of Experts

Scaling without paying for it: experts & routing, top-k gating, the collapse problem & load-balancing loss, capacity/token-dropping, Switch, Mixtral & DeepSeek, expert parallelism.

10 chapters9+ sims
NEW

Test-Time Compute

Letting the model think longer: self-consistency, best-of-N & verifiers, chain-of-thought as compute, tree search, o1-style reasoning RL, the train-vs-test tradeoff, and the overthinking trap.

10 chapters9+ sims
NEW

U-Net

Seeing every pixel: encoder-decoder, the bottleneck blur, skip connections (the one big idea), upsampling & checkerboards, Dice loss, and why it became the backbone of diffusion models.

10 chapters9+ sims
NEW

Linear Attention & RWKV

Escaping the quadratic wall: where the n² comes from, the kernel/associativity trick, the recurrent dual form, the recall catch, RWKV’s decay & gating, the Mamba/RetNet family, and hybrid models.

10 chapters9+ sims
NEW

xLSTM

The LSTM reborn: the two fatal flaws, saturating sigmoids, exponential gating + normalizer (sLSTM), matrix memory (mLSTM), parallelizability, the revision lab, the block, and the recurrent-revival family.

10 chapters9+ sims
NEW

RetNet

The impossible triangle: retention (attention minus softmax plus decay), and one mechanism computed three equivalent ways — parallel to train, recurrent to deploy, chunkwise to scale. Multi-scale heads & the family.

10 chapters9+ sims
NEW

Hyena & Long Convolutions

Attention by convolution: a sequence-length filter, made practical by implicit filters (any length, few params), the FFT (n log n), and data-controlled gating — plus the recurrence=convolution duality.

10 chapters9+ sims
NEW

Jamba

The hybrid that works: interleave mostly-Mamba layers with a few attention layers (for recall) plus MoE (for cheap capacity) — why a few attention layers suffice, the KV-cache memory win, and the hybrid era.

10 chapters9+ sims
NEW

Griffin & Hawk

Google DeepMind’s RNN-speed transformers: a gated linear recurrence (RG-LRU) for cheap global memory + local attention for sharp recall. Linear recurrence → parallel scan → stability → gating → the hybrid stack → constant-size inference.

10 chapters9+ sims
NEW

JEPA

Yann LeCun’s Joint Embedding Predictive Architecture: stop predicting pixels, predict meaning. Why pixel loss blurs → latent prediction → collapse & the EMA fix → predictor + mask tokens → masking → the full I-JEPA loop → world models.

10 chapters9+ sims
NEW

Perceiver & Perceiver IO

One architecture for any modality: funnel a huge input through a small latent bottleneck with cross-attention. The n² wall → latent workspace → cross-attention → latent thinking → Fourier position → iteration → arbitrary outputs via query arrays.

10 chapters9+ sims
NEW

Neural ODEs

Networks with infinite, continuous depth. A residual block is one Euler step — take the limit and depth becomes a flow field solved by an ODE solver: adaptive computation, O(1)-memory adjoint gradients, and the continuous-normalizing-flow ancestor of diffusion.

10 chapters9+ sims
NEW

KANs

Kolmogorov–Arnold Networks: flip the MLP — put learnable 1-D functions (splines) on the edges, let nodes just sum. The result is a network you can read, prune, refine, and turn into a symbolic formula.

10 chapters9+ sims
NEW

Liquid Neural Networks

MIT’s tiny, robust brains: continuous-time neurons whose response speed adapts to the input. Leaky neuron → liquid time constant → bounded stability → robustness → neural circuit policies → closed-form CfC. 19 neurons can drive a car.

10 chapters9+ sims
NEW

Whisper & Audio Transformers

How machines learned to hear: turn sound into a log-mel picture, run a vanilla transformer encoder-decoder over it, control the task with prompt tokens, and train on 680,000 hours of the messy internet. The audio front-end behind all modern speech models.

10 chapters9+ sims
NEW

Video Generation

Sora-style spacetime diffusion: compress video into a spacetime latent, chop it into patches spanning space and time, and denoise the whole clip with a diffusion transformer whose attention crosses frames — coherence by construction, plus emergent world-model behavior.

10 chapters9+ sims
NEW

Diffusion Policy

The denoising that paints images can drive a robot. Instead of regressing one action (and averaging multimodal demos into a crash), it generates an action chunk from noise — committing to a mode, conditioned on observations, replanned in a closed loop. The action head behind modern VLAs.

10 chapters9+ sims
Audio & Speech5
Modalities & Methods5
LLM Inference & Adaptation3
Training Foundations18
NEW

Loss Functions

MSE, cross-entropy, KL divergence, Huber, contrastive, triplet, InfoNCE — every loss derived from scratch with interactive comparisons.

10 chapters12+ sims
NEW

Optimizers

SGD, momentum, AdaGrad, RMSProp, Adam, AdamW, Lion, Sophia — every update rule derived, hand-computed, and raced head-to-head.

10 chapters14+ sims
NEW

Normalization

BatchNorm, LayerNorm, RMSNorm, GroupNorm, InstanceNorm, Pre-LN vs Post-LN — every variant derived, hand-computed, and raced in the Arena.

10 chapters12+ sims
NEW

Attention Variants

MHA, MQA, GQA, sliding window, linear attention, FlashAttention — every variant derived with memory arithmetic and raced in the Arena.

11 chapters14+ sims
NEW

Positional Encoding

Sinusoidal, learned, RoPE, ALiBi, NTK scaling — why transformers need position and how rotation won, with length extrapolation arena.

10 chapters12+ sims
NEW

Activation Functions

Sigmoid, ReLU, GELU, SiLU/Swish, SwiGLU, Mish — every nonlinearity derived with dead neuron analysis and racing arena.

10 chapters12+ sims
NEW

Learning Rate Schedules

Warmup, step decay, cosine annealing, 1cycle, WSD — every schedule derived with loss landscape simulations and racing arena.

10 chapters12+ sims
NEW

Initialization

Xavier, He/Kaiming, orthogonal, transformer recipes — every method derived from variance preservation with deep network explorer.

10 chapters12+ sims
NEW

Gradient Flow

Vanishing/exploding gradients, clipping, accumulation, mixed precision, loss scaling, checkpointing — the complete stability toolkit.

10 chapters12+ sims
NEW

Skip Connections

ResNet residuals, Pre-LN vs Post-LN, DenseNet, Highway — why every modern architecture uses shortcuts and how to choose.

10 chapters12+ sims
NEW

Pooling & Aggregation

Max, average, GAP, CLS, mean, attention, GeM pooling — how to collapse features into fixed-size vectors for any task.

10 chapters12+ sims
NEW

Embedding Layers

Token, position, segment, patch embeddings — tied weights, scaling, subword effects, and the lookup tables behind every model.

10 chapters12+ sims
NEW

Training Loop Mechanics

Epochs, batches, DataLoaders, shuffling, the 6-step training loop, eval mode, common bugs — the complete anatomy of training.

10 chapters12+ sims
NEW

Data Augmentation

Flips, crops, color jitter, RandAugment, Mixup, CutMix, text augmentation, TTA — synthetic variations that prevent overfitting.

10 chapters12+ sims
NEW

Curriculum Learning

Difficulty scoring, Bengio’s classical curriculum, pacing functions, self-paced learning, teacher-student bandits, DoReMi data mixing — and when ordering doesn’t help.

10 chapters9+ sims
NEW

Contrastive Learning

Self-supervised vision without labels: InfoNCE, temperature, the projection head you throw away, MoCo’s queue, and how BYOL & DINO dodge collapse with no negatives.

10 chapters9+ sims
NEW

Knowledge Distillation

Teaching a small model to think like a giant: dark knowledge, temperature softening, the KD loss, feature & attention matching, self-distillation, DistilBERT, and the capacity-gap trap.

10 chapters9+ sims
NEW

Dropout Variants

Co-adaptation, the mask & inverted-dropout scaling, the ensemble view, spatial dropout, DropConnect, stochastic depth/DropPath, DropBlock — breaking your network on purpose to make it generalize.

10 chapters9+ sims
AI & Search2
Mathematics Foundations6
Monte Carlo Methods7
NEW

Monte Carlo Simulation

Estimate the impossible by random sampling: the law of large numbers, the unavoidable square-root error law, why it shrugs off high dimensions, variance reduction — and how it prices options, traces light, and powers modern AI.

12 chapters12 sims
NEW

Importance Sampling

Sample from a proposal you can afford, then reweight to the truth — and watch a one-in-a-million event become reachable. The optimal proposal, effective sample size, rare-event tilting, off-policy RL, and importance-sampled rendering.

11 chapters11 sims
NEW

Quasi-Monte Carlo

Stop scattering points at random — place them deterministically to fill space evenly, and watch integration error fall from one-over-root-N toward one-over-N. Discrepancy, van der Corput, Halton & Sobol, Koksma–Hlawka, effective dimension, and randomized QMC.

12 chapters12 sims
NEW

Markov Chain Monte Carlo

When you can’t sample a distribution, build a random walk that visits its points in the right proportions — then just average. Metropolis–Hastings, Gibbs, detailed balance, burn-in & mixing, autocorrelation & R-hat, and the Ising lattice where it was born.

12 chapters12 sims
NEW

Hamiltonian & Langevin Monte Carlo

Stop wandering — use the gradient of the log-probability to glide across high-dimensional distributions in long, high-acceptance jumps. The score, momentum & leapfrog, NUTS, Langevin & MALA — and how score-based diffusion models are Langevin sampling in disguise.

13 chapters13 sims
NEW

Sequential Monte Carlo

When one proposal isn’t enough: march a weighted swarm of particles through a ladder of distributions, resampling to survive — and read off the normalizing constant nobody else can. Degeneracy & resampling, tempering, annealed importance sampling, and Bayesian model evidence.

12 chapters12 sims
NEW

Monte Carlo Tree Search

When you can’t evaluate every move, search only the promising ones — grow a tree by playing the game out at random, then let the wins steer where you look next. The four phases, UCT exploration, AlphaGo → AlphaZero → MuZero, and LLM reasoning as search (Tree of Thoughts, rStar-Math).

12 chapters12 sims
Evaluation & Analytics9
NEW

Summarizing Distributions

The average salary is $68k and almost nobody earns near it. What a percentile actually is, mean vs median, IQR, the CDF/ECDF, box & violin plots, and why heavy tails make the average lie — descriptive statistics from absolute zero.

11 chapters9 sims
NEW

The Statistics of Evaluation

Is that benchmark delta real or a coin flip? Standard errors, confidence intervals, the bootstrap, McNemar, power, and multiplicity — the statistics that separate signal from sampling noise.

11 chapters9 sims
NEW

Metric Design

A 99%-accurate fraud detector that catches no fraud is a metric you designed badly. Confusion matrices, ROC/PR, calibration, NDCG, aggregation traps, and Goodhart’s law.

11 chapters9 sims
NEW

Plots That Tell the Truth

The same numbers can be plotted into opposite conclusions. Histogram traps, honest error bars, log-log scaling fits, and a lie-detector toolkit for every eval plot you read or make.

11 chapters9 sims
NEW

Experiment Design & A/B Testing

Your feature ‘improved’ retention — because it launched before a holiday. Randomization, power & MDE, the peeking problem, CUPED, switchbacks, and why offline wins vanish online.

11 chapters9 sims
NEW

Regression Testing & Quality Gates

Your tests are green and your model is broken. Golden sets, statistical release gates, PSI/KS drift alarms, canary & shadow, and the rollback machinery that stops bad models from shipping.

11 chapters9 sims
NEW

Evaluating Estimators, Fusion & Robots

Two Kalman filters, identical RMSE — one is lying about its confidence. NEES & NIS consistency, the Cramér–Rao bound, SLAM ATE/RPE, and honest robot success-rate intervals.

11 chapters9 sims
NEW

Evaluating Generative Models

Classification has an answer key; generation has a taste problem. pass@k, Bradley–Terry & Elo, FID from scratch, judge-bias quantified, and inter-rater kappa — for LLMs, VLMs, and diffusion.

11 chapters9 sims
NEW

The Metrics Ladder

Your model gained 2 points and revenue dropped. Climb from module metric to P&L: driver trees, guardrails & SLOs, funnels, error analysis, and exec-ready launch memos.

11 chapters9 sims
Classical Machine Learning11
NEW

Linear Regression & LMS

The first ML algorithm done right. Fit a line, score it with squared error, then minimize two ways: roll downhill with the LMS rule (batch & SGD) and jump to the exact answer with the normal equations — and discover least squares is maximum likelihood under Gaussian noise.

11 chapters9+ sims
NEW

Logistic Regression & Perceptron

How a line learns to say yes/no. Squash a score into a probability with the sigmoid, derive cross-entropy from maximum likelihood, meet the update rule that's identical to LMS, then extend to the perceptron, softmax for many classes, and Newton's method.

10 chapters8+ sims
NEW

Locally-Weighted Logistic Regression

Re-fit a fresh classifier around every query point, trusting nearby data more. The Gaussian kernel, Newton's method, and the bandwidth as a bias/variance knob — the showcase Code Lab repaints the decision boundary as you turn it.

6 chaptersCode Lab
NEW

Generalized Linear Models

One recipe that produces linear regression, logistic regression, softmax, and Poisson regression. The exponential family, why the sigmoid is forced not chosen, the three-line GLM recipe, and the single gradient every GLM shares. The framework that explains the mystery.

10 chapters7+ sims
NEW

Generative Learning: GDA & Naive Bayes

The other half of classification: model what each class looks like, then flip with Bayes. Gaussian Discriminant Analysis (its posterior is secretly a sigmoid), generative vs discriminative trade-offs, and the Naive Bayes spam filter that defined an era — plus Laplace smoothing.

10 chapters7+ sims
NEW

Bias, Variance & Double Descent

Why a model that fits perfectly can predict garbage — and why making it even bigger can fix it. Test error decomposed into bias, variance, and noise; the classical U-curve and its sweet spot; then double descent and the implicit regularization behind modern deep learning, simulated live in your browser.

10 chapters8+ sims
NEW

Model Selection & Regularization

How to actually find the sweet spot without cheating. The golden rule (train/validation/test), the cross-validation family (hold-out, k-fold, leave-one-out), regularization as a continuous complexity dial, L1 vs L2 (shrink vs select), and the Bayesian view where regularization is a prior.

10 chapters9+ sims
NEW

k-Means Clustering

Finding hidden groups when nobody labeled them — the first unsupervised algorithm. The assign/average two-step dance, the distortion it minimizes, why it always converges (and to a local minimum), choosing k with the elbow, and where its spherical assumption breaks.

9 chapters8+ sims
NEW

EM & Gaussian Mixture Models

k-means, leveled up: soft probabilistic clustering with stretchable elliptical groups. The mixture model, the chicken-and-egg problem, the E-step (Bayes’ responsibilities) and M-step (weighted re-fit), why it converges, and how k-means is just EM with hard assignments and fixed spheres.

9 chapters7+ sims
NEW

Principal Component Analysis

Squeeze high-dimensional data down to the few directions that matter, via one eigenvector calculation. Centering, the max-variance projection, the covariance matrix, principal components as its eigenvectors, reduction & reconstruction, and applications from eigenfaces to noise removal.

9 chapters7+ sims
NEW

Independent Component Analysis

Unmix what was blended — the cocktail party problem. The mixing model x=As, finding the unmixing W, independence as the criterion, why ICA needs non-Gaussian sources, a live un-mixing lab, and how ICA differs from PCA (independence beats uncorrelatedness).

9 chapters7+ sims
Language Modeling from Scratch17
Lec 1

Overview & Tokenization

From-scratch LM motivation, the CS336 pipeline, and building BPE tokenizers from character/byte/word up.

10 chapters7 sims
Lec 2

PyTorch & Resource Accounting

Tensors, fp32/bf16/fp8, FLOP counting, the 6ND rule, optimizer & activation memory, the training loop.

10 chapters6 calculators
Lec 3

Architectures & Hyperparameters

Pre-norm, RMSNorm, SwiGLU, RoPE, GQA, FFN ratios — every modern LLM architecture choice, derived.

10 chapters5 canvases
Lec 4

Mixture of Experts

Mixture of Experts: top-k gating, load-balance loss, expert capacity, router z-loss. DeepSeek, Mixtral, Llama 4.

10 chapters5 canvases
Lec 5

GPUs

Arithmetic intensity, the roofline model, memory coalescing, tiling, and fusion — why GPUs idle and how to fix it.

10 chapters5 canvases
Lec 6

Kernels & Triton

HBM traffic, kernel fusion, writing a fused Triton kernel, and FlashAttention's online-softmax recurrence.

10 chapters5 canvases
Lec 7

Parallelism I: Data & Tensor

Collective ops, data parallelism, ZeRO 1–3, ring all-reduce, tensor parallelism, and memory accounting.

10 chapters5 canvases
Lec 8

Parallelism II: Pipeline & FSDP

Pipeline parallelism, the bubble formula, GPipe vs 1F1B, FSDP/ZeRO-3, and sequence parallelism.

10 chapters5 canvases
Lec 9

Scaling Laws I: The Basics

Compute-optimal scaling: C=6ND, the loss law, IsoFLOP curves, Chinchilla, and the D=20N rule.

10 chapters5 canvases
Lec 10

Inference

Prefill vs decode, KV-cache sizing, the roofline, KV compression, speculative decoding, and quantization.

10 chapters5 canvases
Lec 11

Scaling Laws II: Details & Methodology

Parametric loss fits, IsoFLOP profiling, the three Chinchilla methods, WSD, muP transfer, data-limited scaling.

10 chapters5 canvases
Lec 12

Evaluation: How Good Is Your Model?

Perplexity, multiple-choice scoring, majority vote, best-of-N, Elo, and contamination across MMLU, GPQA, Arena.

10 chapters5 canvases
Lec 13

Data I: Sources & The Pipeline

Where LLM corpora come from: Common Crawl, WARC vs WET, HTML→text extraction, the data-mix problem, copyright.

10 chapters5 canvases
Lec 14

Data II: Filtering & Deduplication

Cleaning web text: KenLM & fastText filtering, DSIR, Bloom filters, MinHash, and LSH deduplication.

10 chapters5 canvases
Lec 15

Alignment I: SFT & RLHF

Post-training: SFT, Bradley-Terry preferences, the RLHF KL objective, PPO, and the DPO loss derived.

10 chapters5 canvases
Lec 16

Alignment II: RLVR & GRPO

RLVR with verifiable rewards: REINFORCE, baselines, GRPO group advantage. DeepSeek-R1, Kimi, Qwen 3.

10 chapters5 canvases
Lec 17

Alignment III: Policy Gradients & the RL Frontier

Policy gradients from scratch: the PG theorem, REINFORCE variance, baselines, GAE, PPO clip, GRPO. Series capstone.

10 chapters5 canvases
Principles of Robot Autonomy15
Lec 1

The See-Think-Act Loop & the Autonomy Stack

The four-module autonomy stack and the perception–action loop that ties it together — map representations, finite-state-machine executives, and a full gridworld stack you can break.

10 chapters6 sims · 3 labs
Lec 2

Rigid-Body Transforms, SE(2) & Wheeled Kinematics

From “turn 30°” to wheel speeds: rotation matrices, SE(2) frames, holonomic vs nonholonomic constraints, and the unicycle & differential-drive models you integrate yourself.

11 chapters11 sims · 3 labs
Lec 3

Motion Planning I: Configuration Space & A*

Turn a robot+obstacles into a point searching a graph: C-space inflation, Dijkstra, admissible heuristics, and A* — run by hand on a grid, then built and flooded yourself.

11 chapters8 sims · 3 labs
Lec 4

Sampling-Based Planning: PRM, RRT & RRT*

When grids explode in high dimensions, sample instead: build a roadmap, grow a rapidly-exploring tree, and rewire it to optimality — you implement nearest, steer, and choose-parent yourself.

11 chapters9 sims · 3 labs
Lec 5

Trajectory Optimization & Differential Flatness

A planner gives waypoints; a robot needs a smooth, timed, dynamically-feasible trajectory. Polynomial trajectories, differential flatness (recover v & ω from the path), and the min-snap QP.

11 chapters9 sims · 3 labs
Lec 6

Trajectory Tracking: PID & LQR

Open-loop drifts; feedback corrects. Error dynamics, P/PI/PID, pole placement, and LQR via the Riccati equation — build the controllers and watch them track a path through a disturbance.

11 chapters8 sims · 3 labs
Lec 7

Robot Sensors: IMU, LiDAR, Point Clouds & ICP

What robots measure and how it goes wrong: encoder odometry, gyro drift, LiDAR scans, line extraction, and Iterative Closest Point — you integrate a biased gyro and align two scans with SVD.

12 chapters8 sims · 3 labs
Lec 8

Camera Models & Calibration

How a camera turns 3D into pixels: the pinhole model, intrinsics K, extrinsics, the full projection P=K[R|t], lens distortion, and calibration — you project a cube and calibrate from correspondences.

11 chapters9 sims · 3 labs
Lec 9

Structure from Motion: Features & RANSAC

Recover 3D from images: Harris corners, descriptor matching, epipolar geometry, RANSAC robust fitting (N=log(1−p)/log(1−wₛ)), triangulation and stereo depth Z=fB/d — built by hand.

12 chapters9 sims · 3 labs
Lec 10

Learning-Based & Semantic Perception

From corners to meaning: template matching, IoU & non-max suppression, convolution as a learned filter, detectors and segmentation — and how a detection becomes an FSM input that stops the robot.

11 chapters8 sims · 3 labs
Lec 11

SLAM: Factor Graphs & Pose-Graph Optimization

Build a map and localize in it at once: EKF-SLAM and its growing correlations, factor graphs, Gauss-Newton on a pose graph, sparsity, and loop closure that snaps drift away — built by hand.

12 chapters10 sims · 3 labs
Lec 12

Occupancy Mapping & Frontier Exploration

Build a free/occupied map from noisy range scans via log-odds and ray-casting, then explore autonomously: detect frontiers, score them by information gain vs cost, and cover the unknown.

11 chapters8 sims · 3 labs
Lec 13

Recursive Estimation: Kalman, EKF & UKF

Fuse a noisy prediction with a noisy measurement: the Bayes filter, Gaussians, the Kalman gain as a trust dial, EKF Jacobians for nonlinear models, EKF-localization, and the unscented transform.

12 chapters7 sims · 3 labs
Lec 14

Particle Filters & Monte Carlo Localization

When the belief is multimodal a Gaussian can't capture, use a cloud of weighted samples: predict, weight, resample. Monte Carlo Localization, the kidnapped robot, and N​eff — built from scratch.

11 chapters10 sims · 3 labs
Lec 15

The Frontier: Imitation, VLAs, 3DGS & World Models

Series capstone. Behavior cloning & the O(εT²) distribution-shift trap, DAgger, diffusion policies, Vision-Language-Action models, 3D Gaussian Splatting, and world models — where the classical stack meets learning.

11 chapters9 sims · 3 labs
OpenClaw — Anatomy of an Agent Gateway11
01

The Gateway

Why one long-lived daemon owns the stateful provider connections, and everything else — operator clients, capability nodes, seven chat surfaces — is a thin client of it.

8 chapters3 sims · 3 labs
02

The Wire Protocol

A typed request/response/event grammar over one WebSocket: the handshake, discovery, idempotent retries, and gap-detecting sequence numbers.

8 chapters2 sims · 3 labs
03

Auth, Pairing & Trust

Capability-based trust: keypair device identity, challenge-nonce signing, role/scope RBAC, pairing approval — and what auth actually protects.

9 chapters1 sim · 3 labs
04

The Agent Loop

The run as a serialized state machine — intake to context to inference to tools to stream to persist — with hooks, timeouts, and four ways to end.

9 chapters2 sims · 3 labs
05

Context & Streaming

What the model sees and how you see it back: layered prompt assembly under a cache boundary, and block-and-preview reply streaming.

9 chapters2 sims · 3 labs
06

The Command Queue

Lane-aware concurrency: per-session FIFO plus global caps, and the steer / followup / collect / interrupt modes for messages that arrive mid-run.

9 chapters2 sims · 3 labs
07

Sessions & Multi-Agent

State partitioning and routing at scale: session keying and isolation, lifecycle and pruning, and the most-specific-wins agent router.

9 chapters2 sims · 3 labs
08

Agent Memory

How a stateless model gets durable memory: a file-backed hierarchy, hybrid vector-plus-keyword search, and a dreaming process that consolidates what matters.

9 chapters2 sims · 3 labs
09

Self-Healing & Reliability

Treat failure as normal: retries with backoff and jitter, model failover chains, context-overflow compaction, and diagnostics that recover on their own.

9 chapters2 sims · 3 labs
10

Tools, Exec-Approvals & Sandboxing

The hard enforcement fence: the five-mode exec ladder, the allowlist, exec-approval flow, and sandbox modes that decide whether a command runs.

9 chapters2 sims · 3 labs
11

Skills, Plugins & Automation

The four extension surfaces: skills loaded on demand, plugins, hooks that fire on events with precedence, and cron jobs on a schedule. The finale.

9 chapters2 sims · 3 labs
Flow Matching & Diffusion Models6
Introduction to Robot Learning25
Lec 1

What Is Robot Learning?

Sequential decisions in the physical world: the closed loop, the compounding-error trap that breaks naive imitation, the robot data problem, the four pillars, and a preview of the whole course.

10 chapters6 sims · 3 labs
Lec 2

Robot Learning: An Overview

The agent–environment loop and the reward hypothesis, a policy as a state→action map, the ladder of methods from imitation to RL, known vs learned worlds, and high-level plans over low-level skills.

9 chapters5 sims · 3 labs
Lec 3

ML / DL Refresher, Part 1

Supervised learning, loss & empirical risk, linear and logistic regression by hand, gradient descent and the learning rate, overfitting, the bias–variance tradeoff, and regularization.

10 chapters5 sims · 3 labs
Lec 4

ML / DL Refresher, Part 2

Why depth and nonlinearity, the MLP and backprop by hand, optimizers from SGD to Adam, CNNs, attention and Transformers, generative models, and knowing when the model is uncertain.

10 chapters8 sims · 3 labs
Lec 5

MDP Basics & Imitation Learning, Part 1

The MDP tuple, policies and trajectories, discounted return and value functions, the Bellman idea, and behavior cloning — plus why copying an expert is not standard supervised learning.

10 chapters5 sims · 3 labs
Lec 6

Imitation Learning, Part 2

DAgger and how it beats the quadratic cost, privileged teachers, GAIL as distribution matching with a discriminator, and Diffusion Policy for multimodal demonstrations.

10 chapters7 sims · 3 labs
Lec 7

RL Basics: Value & Policy Iteration

Value and Q functions, the Bellman expectation and optimality equations derived, dynamic programming — policy evaluation, policy improvement, policy iteration and value iteration — worked by hand on a gridworld.

10 chapters6 sims · 3 labs
Lec 8

Q-Learning & Variants

Model-free control from samples: Monte Carlo vs TD, SARSA vs Q-learning and the off-policy max, ε-greedy exploration, the cliff, and DQN's replay buffer and target network that tame the deadly triad.

10 chapters7 sims · 3 labs
Lec 9

Policy Gradient Methods

Optimize the policy directly: the log-derivative trick, the policy gradient theorem, REINFORCE, why the estimator is unbiased but high-variance, and reward-to-go and baselines that calm it down.

10 chapters5 sims · 3 labs
Lec 10

Actor-Critic Methods

A learned critic cuts policy-gradient variance: the advantage function, the TD error, two networks in one loop, and the n-step / GAE dial that trades bias against variance.

10 chapters6 sims · 3 labs
Lec 11

Advanced RL: TRPO, PPO, DDPG & SAC

The biggest safe step: natural gradients and the KL trust region, PPO's clipped surrogate, deterministic off-policy control with DDPG/TD3, and maximum-entropy RL with SAC — mapped on one 2×2 grid.

10 chapters6 sims · 3 labs
Lec 12

Model-Based Control Basics

When you know the physics: state-space models, feedback, PID, stability from eigenvalues, and LQR — the optimal linear controller — with the Riccati recursion worked by hand.

10 chapters5 sims · 3 labs
Lec 13

Optimal Control & Planning, Part 1

Plan a whole control sequence: shooting vs collocation, the LQR backward pass derived line by line, and iLQR / DDP — linearize, solve LQR, repeat — plus sequential convex programming.

10 chapters5 sims · 3 labs
Lec 14

Optimal Control & Planning, Part 2

When you can't write the equations: MPC and replanning, random shooting, the cross-entropy method and MPPI, and learned-model RL — PILCO, PETS ensembles, and MBPO.

11 chapters5 sims · 3 labs
Lec 15

Deep Model-Based RL: Dreamer & TD-MPC

Learn a latent world model from pixels and plan inside it: Dreamer's learning-in-imagination, TD-MPC's plan-in-latent with a learned value, and how model error compounds over a rollout.

10 chapters3 sims · 3 labs
Lec 16

Learning Structured World Models

Guest lecture: model dough, beans, and fluids as particles plus relations and learn the dynamics with a graph network — message passing, permutation-equivariance, and planning with a learned simulator.

11 chapters3 sims · 3 labs
Lec 17

Offline Reinforcement Learning

Guest lecture: learn from a fixed log with no new interaction — the out-of-distribution overestimation death spiral, and the fixes: policy constraints, conservatism (CQL), in-sample IQL, and Diffuser.

10 chapters6 sims · 3 labs
Lec 18

Inverse Reinforcement Learning

Recover the reward from demonstrations: why it's ambiguous, feature matching, max-margin, and the maximum-entropy principle — the least-committal reward that explains the expert.

10 chapters6 sims · 3 labs
Lec 19

Bandits & Preference-Based Learning

The cleanest explore–exploit problem: regret, ε-greedy, UCB optimism, Thompson sampling, and dueling bandits / preference learning — the foundation under RLHF.

11 chapters5 sims · 3 labs
Lec 20

Exploration in Reinforcement Learning

When rewards are sparse and random actions never stumble onto them: count-based bonuses, curiosity as prediction error, the noisy-TV trap, and Random Network Distillation.

11 chapters3 sims · 3 labs
Lec 21

Robot Simulation & Sim2Real

The reality gap: why a sim-perfect policy falls on the real robot, domain randomization (reality is just sim #10,001), system identification, and champion-level drone racing.

10 chapters5 sims · 3 labs
Lec 22

Safe RL & Safe Robot Learning

When you can't learn from the crash: constrained MDPs, safe exploration, control barrier functions, and safety filters that project any unsafe action onto the nearest safe one.

10 chapters3 sims · 3 labs
Lec 23

Multi-Task, Adaptive & Transferable Learning

Adapt within a step: teacher–student distillation, RMA inferring terrain from recent motion, and Neural-Fly's online residual adaptation for agile flight in wind.

11 chapters4 sims · 3 labs
Lec 24

Foundation Models in Robotics

Guest lecture: borrow the web's common sense — SayCan's say×can grounding, CLIPort, RT-1's tokenized actions, Code-as-Policies, and the move toward vision-language-action models.

10 chapters5 sims · 3 labs
Lec 25

Course Summary: The Robot-Learning Spine

The capstone: the two hard problems revisited, one map from imitation to foundation models, a method-selection decision tree, honest tradeoffs, and the open frontier.

10 chapters7 sims · 3 labs
Sensor Fusion: Classical to Modern22
Lec 1

The Fusion Problem: Why Fuse At All

Why no single sensor is enough: complementary vs competitive vs cooperative fusion, the fusion levels, and the reliability paradox — more sensors isn't always more robust.

11 chapters7 sims · 3 labs
Lec 2

Fusion Architectures & the JDL Model

Where and how to combine: the JDL levels, centralized vs distributed vs decentralized, and the double-counting trap — correlated estimates that fool the fused covariance into overconfidence.

11 chapters7 sims · 3 labs
Lec 3

Combining Two Measurements

The seed of all fusion: the inverse-variance weighted average. Trust each source by its precision; the fused estimate is always sharper — and it's secretly the Kalman update.

11 chapters7 sims · 3 labs
Lec 4

The Bayes Filter

Fusion over time: keep a belief, predict it forward (it blurs), then fuse each new measurement via Bayes' rule (it sharpens). The recursive engine behind every filter that follows.

11 chapters6 sims · 3 labs
Lec 5

The Kalman Filter as Optimal Fusion

The Bayes filter for linear-Gaussian worlds: fuse a model prediction with a measurement, each weighted by its uncertainty. The Kalman gain is the trust dial — recursive inverse-variance fusion.

11 chapters6 sims · 3 labs
Lec 6

Complementary Filters & AHRS

The cheapest fusion: blend a fast-but-drifty gyro with a slow-but-stable accel/mag by frequency — high-pass the gyro, low-pass the reference. Madgwick & Mahony attitude estimators.

11 chapters6 sims · 3 labs
Lec 7

The Extended Kalman Filter

Nonlinear fusion: linearize the model with its Jacobian, then run the Kalman machinery — fusing IMU motion with range/bearing GPS. The workhorse of real navigation.

11 chapters6 sims · 3 labs
Lec 8

The Unscented Kalman Filter

Skip the Jacobian: pick a few sigma points, push them through the true nonlinearity, and recover the fused mean and covariance. More accurate than the EKF, derivative-free.

11 chapters5 sims · 3 labs
Lec 9

The Information Filter

The Kalman filter inside-out: track information (inverse covariance) instead of covariance. Fusion becomes simple addition — the key that unlocks decentralized, multi-sensor networks.

11 chapters7 sims · 3 labs
Lec 10

The Particle Filter

When the world isn't Gaussian: represent the belief with a swarm of weighted samples. Predict, weight by the measurement, resample — non-parametric fusion that powers Monte Carlo Localization.

11 chapters6 sims · 3 labs
Lec 11

Covariance Intersection

Fusing estimates that secretly share information: when the correlation is unknown, naive fusion grows overconfident and diverges. CI stays consistent for any correlation — the key to decentralized fusion.

11 chapters7 sims · 3 labs
Lec 12

Data Association

The dodged half of fusion: which measurement updates which track? Validation gating, global nearest-neighbor, JPDA and MHT — because one wrong correspondence makes the best filter fuse garbage and diverge.

12 chapters6 sims · 3 labs
Lec 13

IMU & Inertial Navigation

The sensor everything fuses with: what gyros and accelerometers really measure (specific force, not acceleration), strapdown mechanization, and why an unaided IMU's position error explodes as t³.

11 chapters5 sims · 3 labs
Lec 14

INS/GNSS Coupling

Bounding inertial drift with absolute fixes: loose vs tight vs deep coupling, and the error-state Kalman filter that fuses a high-rate IMU with low-rate GNSS — coasting through tunnels, snapping back on reacquisition.

11 chapters7 sims · 3 labs
Lec 15

Calibration & Time Sync

The silent prerequisite for all fusion: get the sensor-to-sensor transform (hand-eye AX=XB) or the clock offset wrong and your optimal filter quietly diverges. Spatial + temporal calibration, observability, Kalibr.

11 chapters7 sims · 3 labs
Lec 16

Visual-Inertial Odometry

The complementary masterpiece: a camera and an IMU cover each other's blind spots — and the IMU's metric acceleration resolves the monocular scale. MSCKF, VINS-Mono, and IMU preintegration.

11 chapters5 sims · 3 labs
Lec 17

LiDAR-Inertial & Factor Graphs

Fusing precise 3-D LiDAR with the IMU that de-skews it — and the factor graph that unifies it all. Point-to-plane registration, LIO-SAM vs FAST-LIO2, iSAM2, and degeneracy detection.

11 chapters5 sims · 3 labs
Lec 18

Multimodal Driving Fusion

Camera, LiDAR, and radar in a self-driving stack: their complementary blind spots, the early/mid/late fusion question, PointPainting, and the Bird's-Eye-View revolution (Lift-Splat-Shoot).

11 chapters6 sims · 3 labs
Lec 19

Transformer & BEV Fusion

When attention learns where to fuse: BEVFormer, BEVFusion, and TransFusion replace hand-designed projection with soft cross-modal attention — gracefully down-weighting a blinded or miscalibrated sensor.

10 chapters6 sims · 3 labs
Lec 20

Learned & Differentiable Filters

The frontier: keep the Bayesian filter's structure but learn the parts you don't know. KalmanNet learns the gain, BackpropKF learns the measurement uncertainty — model-based deep learning.

11 chapters7 sims · 3 labs
Lec 21

Debugging & Consistency

The capstone skill: how to know your fusion works. NEES/NIS chi-square tests, innovation whiteness, and robust costs (Huber, DCS, switchable constraints) that catch the silent killer — overconfidence.

12 chapters5 sims · 3 labs
Atlas

The Sensor Fusion Atlas

The ultimate map: an interactive concept graph wiring all 21 methods together, a 1960–2024 historical timeline, and the full curated reference library. One page for the whole journey.

connections · timeline · refsinteractive map
Visual Navigation for Autonomous Vehicles14
Lec 1

3D Geometry: Rotations & Transforms

Frames, rotation matrices SO(3), Euler angles & gimbal lock, axis-angle, quaternions, and SE(3) transforms.

10 chapters5 canvases
Lec 4–5

Lie Groups: SO(3), SE(3) & exp

Lie groups for optimizing over rotations: so(3), hat/vee, the exp & log maps, the SE(3) Jacobian, BCH, retraction.

10 chapters6 canvases
Lec 6–7

Control: PID, LQR & Geometric

Control: state-space stability, PID, LQR via Riccati, quadrotor dynamics, and geometric attitude control on SO(3).

10 chapters5 canvases
Lec 8–10

Trajectory Optimization: Minimum-Snap & Differential Flatness

Minimum-snap trajectories: smoothness as derivative minimization, the QP formulation, and differential flatness.

10 chapters5 canvases
Lec 11

Image Formation: Pinhole Camera Model

Pinhole projection (x=fX/Z), intrinsic matrix K, extrinsics [R|t], full pipeline P=K[R|t], back-projection, and lens distortion.

10 chapters5 canvases
Lec 12–13

Feature Detection & Tracking

Corners vs edges, structure tensor G, Harris & Shi-Tomasi, SIFT descriptors, ratio-test matching, KLT optical flow.

10 chapters5 canvases
Lec 14

Two-View Geometry: Epipolar & SfM

Epipolar constraint, essential matrix E=[t]×R, fundamental matrix F, 8-point algorithm, 4-fold R,t ambiguity, cheirality, triangulation, scale ambiguity.

10 chapters5 canvases
Lec 15

RANSAC: Robust Estimation Amid Outliers

Minimal sample → fit → count inliers → keep best → refit; iteration formula N=log(1−p)/log(1−ws); threshold tradeoffs; sample sizes line=2/homography=4/E=8.

10 chapters5 canvases
Lec 16–18

Estimation & Nonlinear Least Squares

MLE → weighted LS; normal equations (ATA)x=ATb; Jacobian as sensitivity matrix; Gauss-Newton linearize-and-step; LM damping λI trust-region; covariance = (JTΩJ)−1.

10 chapters5 canvases
Lec 18–19

Optimization on Manifolds

SO(3)/SE(3) as curved surfaces; tangent space; retraction R·exp(δ); ⊞/⊝ operators; on-manifold GN: perturb in tangent → solve → retract; rotation averaging; singularity-free local coords.

10 chapters5 canvases
Lec 20

Visual-Inertial Odometry (VO/VIO)

VO pipeline: track → relative pose → chain SE(3); drift & why it compounds; monocular scale ambiguity; IMU accelerometer & gyro; bias double-integration; VIO fusion; IMU preintegration ΔR/v/p; tight vs loose coupling.

10 chapters5 canvases
Lec 21–22

Place Recognition: Bag-of-Words & Loop Closure

Image retrieval framing; visual vocabulary via k-means; BoW histogram; TF-IDF weighting; vocab tree + inverted index; cosine similarity ranking; geometric verification (RANSAC); precision-over-recall; perceptual aliasing; DBoW2/FAB-MAP.

10 chapters5 canvases
Lec 23–24

SLAM: Factor Graphs & Sparsity

Factor graph = poses + landmarks + factors; MAP → sparse NLLS; odometry, landmark, loop-closure factors; sparsity from local measurements; Hessian block pattern; sparse Cholesky & variable ordering; bundle adjustment; smoothing vs filtering; marginalization & Schur complement fill-in.

10 chapters5 canvases
Lec 29–30

Robust SLAM & Frontiers

M-estimators (Huber, Cauchy, TLS), influence functions, IRLS, non-convexity & local minima, GNC annealing, switchable constraints, PCM, SE-Sync certifiability; frontiers: NeRF/Gaussian maps, learned features, semantic SLAM, multi-robot, LiDAR, event cameras; full VNAV series cheat sheet.

10 chapters5 canvases
CAPSTONE

The Estimation–Learning Merger

Where the two halves meet: differentiable filters, neural nets inside Kalman skeletons, KalmanNet's learned gain, transformers that filter in their forward pass, and FoundationStereo front-ends on factor-graph back-ends — gated by chi-squared consistency.

10 chapters9+ sims
FIELD GUIDE

State, Doubt & Learning Machines

The whole arc — Kalman filters to robot foundation models — plus a build ladder of 8 runnable in-browser Python labs: implement a Kalman filter, EKF Jacobian, unscented transform, particle filter, pose-graph SLAM, VIO bias estimation, Q-learning vs SARSA, and a diffusion policy yourself.

4 chapters8 code labs
TinyML & Efficient Deep Learning17
Lec 21

On-Device Training & Transfer Learning

On-device training: why activations explode, TinyTL bias-only, sparse backprop, gradient checkpointing, federated learning. Finale.

10 chapters5 canvases
Lec 19–20

Distributed Training: Data, Tensor & Pipeline Parallelism

Distributed training: the memory wall, ring all-reduce, ZeRO stages, pipeline bubbles, tensor & 3D parallelism.

10 chapters5 canvases
Lec 15

Long-Context LLMs: StreamingLLM & KV-Cache Compression

Long context: the KV crossover, position interpolation, attention sinks, StreamingLLM, H2O, INT4 KV, DuoAttention.

10 chapters5 canvases
Lec 14

Efficient LLM Post-Training: LoRA, QLoRA & PEFT

Parameter-efficient fine-tuning: LoRA, QLoRA with NF4, double quantization, and multi-adapter serving.

10 chapters5 canvases
Lec 13

LLM Deployment & Serving in Practice

Serving LLMs: SmoothQuant, AWQ, GPTQ, W4A16, PagedAttention, continuous batching, speculative decoding.

10 chapters5 canvases
Lec 12

Efficient Transformers & LLMs

Efficient attention: O(N²) cost, KV-cache mechanics, MQA/GQA, sparse/linear attention, RoPE, ALiBi.

10 chapters5 canvases
Lec 11

TinyEngine: Inference Systems for Tiny Devices

TinyEngine: loop reordering, tiling, unrolling/SIMD, im2col vs direct conv, in-place depthwise, fusion, codegen.

10 chapters5 canvases
Lec 10

MCUNet: Deep Learning on Microcontrollers

MCUNet: the SRAM vs Flash bottleneck, TinyNAS co-design, and patch-based inference on a 256 KB MCU.

10 chapters5 canvases
Lec 9

Knowledge Distillation: Teaching Small Networks

Knowledge distillation: dark knowledge, temperature, the KD loss, feature/attention matching, self-distillation.

10 chapters4 canvases
Lec 8

NAS II: Hardware-Aware & OFA

Hardware-aware NAS: latency tables, ProxylessNAS differentiable latency, and Once-for-All progressive shrinking.

10 chapters5 canvases
Lec 7

NAS I: Searching Architectures

Neural architecture search: search spaces, RL & evolutionary strategies, DARTS, and weight-sharing supernets.

10 chapters5 canvases
Lec 1

Why Efficiency & Metrics

The edge-hardware gap and the metrics that bridge it: model size, MACs vs FLOPs, activation memory, energy, roofline.

10 chapters5 canvases
Lec 2

NN Building Blocks & Compute

Exact params & MACs for Linear, Conv, depthwise-separable, BatchNorm folding, and the Transformer block.

10 chapters5 canvases
Lec 3

Pruning I: Sparsity

Pruning: the formal problem, granularities, importance criteria (magnitude, BN, OBD), iterative prune–finetune, 2:4 sparsity.

10 chapters7 canvases
Lec 4

Pruning II: Lottery & HW

Lottery Ticket Hypothesis, IMP rewinding, AMC & NetAdapt, CSR encoding, EIE, and Ampere 2:4 sparse cores.

10 chapters5 canvases
Lec 5

Quantization I: Fewer Bits

Number formats, K-means weight clustering (Deep Compression), and linear quantization r=S(q−Z) derived.

10 chapters5 canvases
Lec 6

Quantization II: PTQ/QAT/LLM

Activation quantization, calibration (min/max, percentile, KL), QAT via the STE, SmoothQuant, GPTQ, AWQ.

10 chapters5 canvases
Signal Processing for ML31
Lec 1

Introduction

SP+ML overview, applications, course roadmap, full pipeline demo.

8 chPilanci
Lec 2

Discrete Signals

x[n], δ[n], periodicity theorem, Nyquist sampling, aliasing.

8 chPilanci
Lec 3

Quantization Noise

Bennett’s theorem, SQNR, 6 dB/bit rule, proof sketch.

8 chPilanci
Lec 4

Lloyd-Max Quantizers

Non-uniform, centroid/boundary conditions, 1/3 power law, NF4 for LLMs.

10 chPilanci
Lec 5

Dithering & Stochastic Rounding

Linearization, subtractive dither, NVFP4, gradient accumulation.

8 chPilanci
Lec 6

DFT

Fourier basis, orthogonality, DFT/IDFT, FFT butterfly.

8 chPilanci
Lec 7

Spectral Descriptors

Centroid, spread, kurtosis, entropy, flatness, flux.

8 chPilanci
Lec 8

STFT

Windowed DFT, spectrogram, uncertainty principle, overlap-add.

8 chPilanci
Lec 9

Distance Classification

NN classifier, Hilbert spaces, Parseval, template matching.

8 chPilanci
Lec 10

Wavelets

CWT, DWT filter bank, Haar, Daubechies, multi-resolution.

8 chPilanci
Lec 11

Wavelet Applications

Denoising, JPEG2000, wavelet families, Fourier vs wavelet.

8 chPilanci
Lec 12

Linear Systems

LTI, convolution theorem, eigenvectors, circulant matrices.

8 chPilanci
Lec 13

Cepstrum & MFCC

Log trick, mel scale, mel filter bank, MFCC pipeline.

8 chPilanci
Lec 14

Bayes Classifiers

Bayes risk, likelihood ratio, ROC curves, Neyman-Pearson.

8 chPilanci
Lec 15

LDA & QDA

Stationarity, autocorrelation, Gaussian classification, LDA/QDA.

8 chPilanci
Lec 16

Fisher Discriminant

J(w) criterion, scatter matrices, simultaneous diagonalization.

8 chPilanci
Lec 17

SVM

Maximum margin, hard/soft margin, slack variables, multi-class.

8 chPilanci
Lec 18

Convex Duality

Lagrangian, KKT, primal→dual derivation, dual SVM.

8 chPilanci
Lec 19

Kernels

Feature maps, kernel trick, RBF, Mercer, representer theorem.

8 chPilanci
Lec 20

Regression & AR

Least squares, ridge, LASSO, autoregressive models, Yule-Walker.

8 chPilanci
Lec 21

RKHS

Bayesian regression, kernel regression, Gaussian processes.

8 chPilanci
Lec 22

Adaptive Filters

Wiener filter, LMS algorithm, noise cancellation, echo cancel.

8 chPilanci
Lec 23

Neural Networks

Hidden layers, activations, backprop, universal approximation.

8 chPilanci
Lec 24

Deep Learning & CNNs

Convolutional layers, spectrograms+CNN, BatchNorm, ResNets.

8 chPilanci
Lec 25

Attention

QKV, scaled dot-product, multi-head, transformer blocks.

8 chPilanci
Lec 26

Transformers for Signals

Signal tokenization, causal masking, GPT-style prediction.

8 chPilanci
Lec 27

Diffusion Models

Forward/reverse process, noise prediction, WaveGrad.

8 chPilanci
Lec 28

Convex Neural Networks

Pilanci’s reformulation, group LASSO, double descent.

8 chPilanci
Lec 29

Autoencoders & RPCA

VAE, nuclear norm, Robust PCA, matrix separation.

8 chPilanci
Lec 30

NMF & Clustering

Multiplicative updates, source separation, deep MF.

8 chPilanci
Lec 31

Dictionary Learning

K-SVD, OMP, matching pursuit, LASSO, sparse coding.

8 chPilanci
Embedded Systems & IoT7
Model Optimization4
Systems & Hardware1
Systems for Machine Learning9

Stanford CS 229s Fall 2023. Hardware-aware algorithm design, transformer efficiency, FlashAttention, sparsity & quantization, finetuning, parallelism, efficient architectures, cluster scheduling.

PyTorch Deep Dive5
Tools & Frameworks1
Probability Distributions5
Estimation & Probabilistic Models12
NEW

Continuous-Time Markov Chains

Discrete chains tick on a clock; real systems jump whenever they like. The generator matrix, exponential clocks racing to jump, the transition function, equilibrium, and Gillespie simulation — built from zero.

12 chapters12 sims
NEW

Continuous-Time HMM

A hidden state that drifts in continuous time, watched only at scattered moments — the gap between observations becomes part of the math. Forward-backward with the matrix exponential, smoothing, and continuous-time EM.

12 chapters12 sims
NEW

Particle Filter

Estimation by a cloud of weighted guesses — for multimodal, nonlinear beliefs the Kalman filter can’t represent. Predict, weight, resample; Monte Carlo Localization; the curse of dimensionality.

10 chapters8+ sims
NEW

Factor Graphs

The modern SLAM back-end: optimize the whole trajectory + map at once as a sparse least-squares problem. Smoothing vs filtering, variables & factors, Gauss-Newton, the loop closure that snaps a drifted map straight, and bundle adjustment.

10 chapters9+ sims
NEW

Robust Estimation

One bad measurement wrecks least squares. The two great cures: down-weight outliers (M-estimators, Huber, IRLS) and vote them out (RANSAC). Influence functions, breakdown points, and the robust kernels behind every SLAM back-end and SfM pipeline.

10 chapters9+ sims
LESSON 14

Bayes Filter

The mother of all filters — recursive belief update from first principles.

9 chapters
LESSON 15

Kalman Filter

Track a moving object through noise — the most elegant algorithm in engineering.

11 chapters
LESSON 16

Extended Kalman Filter

When the world is nonlinear — linearize with Jacobians.

9 chapters
LESSON 17

Unscented Kalman Filter

Beyond linearization — sigma points capture nonlinearity directly.

9 chapters
LESSON 18

Hidden Markov Model

Sequences with hidden causes — Forward, Viterbi, Baum-Welch.

10 chapters
LESSON 19

Bayes Estimation

Prior × likelihood = posterior. The foundation of all inference.

9 chapters
LESSON 20

Bayesian Networks

Variables with dependencies — DAGs, d-separation, message passing.

10 chapters
Robotics & Perception5
Decision & Control3
Deep Reinforcement Learning18
43

Policy Gradients: The Complete Guide

Every symbol explained, every step derived. From REINFORCE to off-policy IS to KL constraints → PPO.

12 chaptersCS 224R
44

Actor-Critic Methods

Learn to estimate what’s good vs bad. MC, bootstrapping, N-step returns, PPO, SAC — the complete guide.

12 chaptersCS 224R
45

Off-Policy Actor-Critic: PPO & SAC

The practical algorithms. Clipping, GAE, replay buffers, Q-functions — from theory to real robots.

11 chaptersCS 224R
47

Offline RL: Learning from Fixed Data

Train policies without environment interaction. AWR, AWAC, IQL — data stitching, expectile regression, and implicit policy improvement.

10 chaptersCS 224R
48

Reward Learning

Where do rewards come from? Goal classifiers, adversarial training, human preferences, Bradley-Terry, RLHF, and Constitutional AI.

10 chaptersCS 224R
46

Q-Learning & DQN

No policy needed. Bellman optimality, target networks, double Q-learning, N-step returns — value-based RL.

11 chaptersCS 224R
47

Imitation Learning

Behavioral cloning, expressive policies, diffusion policies, compounding errors, DAgger — learning from demos.

11 chaptersCS 224R
48

Deep RL: The Complete Theory

Every derivation, every proof. MDPs → Policy Gradients → TRPO → PPO → SAC → DQN → Offline RL → DPO.

12 chaptersCS 224R · CS 234 · CS 285
49

Model-Based RL

Learn a simulator of the world, then practice inside it. Dyna, MBPO, MPC, CEM, value-augmented planning, AlphaGo as MBRL.

10 chaptersCS 224R
50

Multi-Task & Goal-Conditioned RL

One policy, many tasks. Task conditioning, goal-conditioned rewards, HER — turning failures into free training data.

10 chaptersCS 224R
51

CS 224R Exam Review

Complete midterm prep: every algorithm, every equation, 36+ practice questions, interactive exam simulator, cheat sheet.

12 chaptersCS 224R
52

RL for LLM Reasoning

Noam Brown’s approach: search, self-play, and RL for superhuman reasoning in language models.

CS 224R
53

RLHF & DPO

The post-training frontier: reward models, PPO alignment, direct preference optimization, and beyond.

CS 224R
54

Meta-Reinforcement Learning

Learning to learn new tasks from a handful of episodes. Black-box architectures, exploration strategies, task inference as POMDP.

12 chaptersCS 224R
55

Hierarchy in IL & RL

Decompose long-horizon tasks into subtask hierarchies. HL/LL policies, goal representations, DAgger for hierarchy, HIRO, skill discovery.

12 chaptersDeep RL
56

Sim2Real Robot Learning

Crossing the reality gap: domain randomization, privileged teacher–student adaptation, real2sim residuals & actuator nets, learning from human data, and the Sim2Real 1.0→4.0 map.

11 sectionsCS 224R
57

RL for Robot Foundation Models

Pushing a vision-language-action model past the 80% imitation plateau with RL: offline-RL-as-supervised-learning, diffusion steering, and small edit policies. When the model is too big, too weird, too expensive for textbook RL.

10 sectionsCS 224R
58

Frontiers of Deep RL & How to Do Research

The capstone: deep RL as a unified toolbox, the four data settings, the seven open frontiers (no reward, prior knowledge, world models, scaling, safety, errors, evaluation) — and how to actually pick problems, front-load risk, and do research.

10 sectionsCS 224R
NLP with Deep Learning19
Lec 1

History of Language AI

Four eras of NLP: rule-based translation, hand-built AI, statistical methods, neural revolution. Understanding vs pattern matching.

10 chaptersCS224N
Lec 2

Word Vectors

One-hot limitations, distributional hypothesis, Word2Vec (CBOW & Skip-gram), GloVe, negative sampling, word analogies, embedding bias.

10 chaptersCS224N
Lec 3

Backpropagation & Neural Nets

Neurons, activations, forward pass, loss functions, chain rule, backprop algorithm, gradient descent playground, practical tips.

10 chaptersCS224N
Lec 4

Language Models & RNNs

N-grams, neural LMs, recurrent networks, BPTT, vanishing gradients, LSTM & GRU, live text generation, attention preview.

10 chaptersCS224N
Lec 5

The Transformer

Self-attention, scaled dot-product, multi-head attention, positional encoding, encoder-decoder, build a Transformer piece by piece.

10 chaptersCS224N
Lec 6

Practical Methodology

Debugging, hyperparameter tuning, regularization, training diagnostics dashboard, evaluation & error analysis.

8 chaptersCS224N
Lec 7

Pre-training at Scale

BERT vs GPT paradigms, masked LM, autoregressive LM, data pipelines, scaling laws, Chinchilla, Llama 3 architecture.

10 chaptersCS224N
Lec 8

Post-training: RLHF & DPO

SFT, reward modeling, PPO with KL constraint, DPO, alignment data, evaluation, safety guardrails.

10 chaptersCS224N
Lec 9

Efficient Adaptation

In-context learning, prompt engineering, chain-of-thought, lottery tickets, adapters, LoRA, PEFT playground.

10 chaptersCS224N
Lec 10

Agents, Tools & RAG

Retrieval-augmented generation, dense retrieval, ReAct reasoning loops, Toolformer, agent simulator.

10 chaptersCS224N
Lec 11

Benchmarking & Evaluation

MMLU, HELM, LLM-as-judge, benchmark explorer dashboard, data contamination, metric saturation.

8 chaptersCS224N
Lec 12

Reasoning Part 1

Chain-of-thought, self-consistency, zero-shot CoT, process rewards, DeepSeek-R1, GRPO & DAPO.

10 chaptersCS224N
Lec 13

Reasoning Part 2

Process reward models, best-of-N, speculative decoding, RoPE, context extension, test-time compute scaling.

10 chaptersCS224N
Lec 14

Tokenization & Multilinguality

BPE, WordPiece, SentencePiece, multilingual fertility, XLM-R cross-lingual transfer, tokenizer explorer.

8 chaptersCS224N
Lec 15

Interpretability

Linear probes, attention visualization, sparse autoencoders, agentic interpretability, concept discovery.

8 chaptersCS224N
Lec 16

Social & Broader Impacts

Bias, toxicity, misinformation, privacy, environmental cost, governance & policy.

8 chaptersCS224N
Lec 17

Multimodality

Vision encoders, early/late fusion, Chameleon, Transfusion, scaling mixed-modal, retrieval-augmented multimodal.

10 chaptersCS224N
Lec 18

LoRA Without Regret

When LoRA fails, regularization in PEFT, merging adapters, QLoRA, DoRA, future of adaptation.

8 chaptersCS224N
Lec 19

Open Questions in NLP 2026

Reasoning, grounding, efficiency, safety, multimodal frontier, interactive research map.

8 chaptersCS224N
Machine Learning with Graphs19
Lec 1

Introduction to Graph ML

Why graphs? Node/edge/graph-level tasks. AlphaFold, PinSage, drug interactions, antibiotic discovery.

10 chaptersCS224W
Lec 2

Node Embeddings

Encoder-decoder framework, DeepWalk, node2vec (biased walks with p,q), negative sampling, matrix factorization view.

10 chaptersCS224W
Lec 3

Graph Neural Networks

Message passing, computation graphs, GCN (mean aggregation), GNNs generalize CNNs, transformers as GNNs.

10 chaptersCS224W
Lec 4

GNN Design Space

Message + aggregation framework. GCN vs GraphSAGE vs GAT. Multi-head attention. Skip connections, over-smoothing.

10 chaptersCS224W
Lec 5

GNN Augmentation & Training

Feature/structure augmentation, virtual nodes, neighbor sampling, prediction heads, loss functions, dataset splitting.

10 chaptersCS224W
Lec 6

Theory of GNNs

How powerful are GNNs? WL test, multiset injectivity, why GCN/GraphSAGE fail, GIN achieves maximum expressiveness.

10 chaptersCS224W
Lec 7

Powerful Graph Encoders

Beyond WL: Laplacian eigenvectors, structural features, position-aware GNNs, anchor sets, higher-order k-WL.

10 chaptersCS224W
Lec 8

Graph Transformers

Self-attention on graphs, positional encodings (Laplacian, random walk), Graphormer, GPS hybrid architecture.

10 chaptersCS224W
Lec 9

Heterogeneous Graphs

Multiple node/edge types. RGCN (relation-specific weights), basis decomposition, HGT, metapaths.

10 chaptersCS224W
Lec 10

Knowledge Graphs

TransE, TransR, DistMult, ComplEx, RotatE. Relation patterns: symmetry, composition, 1-to-N. KG completion.

10 chaptersCS224W
Lec 11

GNNs for RecSys

Bipartite graphs, NGCF, LightGCN, BPR loss, PinSage scalability, cold start, GNN vs LLM.

10 chaptersCS224W
Lec 12

Relational Deep Learning

Databases ARE graphs. Foreign keys as edges, GNNs on relational data, RelBench, vs XGBoost.

10 chaptersCS224W
Lec 13

Advanced RDL

Temporal message passing, multi-hop aggregation, Griffin universal encoder, schema-specific vs universal.

10 chaptersCS224W
Lec 14

Advanced GNN Topics

Graph foundation models, pre-training, explainability, equivariant GNNs, dynamic graphs, practical tips.

10 chaptersCS224W
Lec 15

KG Foundation Models

Inductive KG reasoning, ULTRA, cross-schema transfer, foundation models for knowledge graphs.

10 chaptersCS224W
Lec 16

LLM + GNN

LLMs as feature encoders, GNNs as structure encoders, joint pipelines, graph-augmented LLMs.

10 chaptersCS224W
Lec 17

Agents + Graphs

ReAct, Reflexion, graph-grounded agents traversing KGs step by step. Tool use, PPO/DPO optimization, STaRK benchmarks.

10 chaptersCS224W
Lec 18

Deep Generative Models for Graphs

GraphRNN two-RNN sequential generation, GCPN RL-guided design, JT-VAE motifs, diffusion on graphs, molecular benchmarks.

10 chaptersCS224W
Lec 19

Conclusion & The Road Ahead

Full arc from node embeddings to graph foundation models. Open problems, research frontiers, practical advice for GNN practitioners.

8 chaptersCS224W
Deep Learning for Computer Vision28
28

Image Classification

k-NN, L1/L2 distance, hyperparameter tuning, cross-validation.

10 chapters
29

Linear Classification

Score function Wx+b, hinge loss, cross-entropy, regularization.

10 chapters
30

Loss Functions

MSE, cross-entropy, KL divergence, Huber, contrastive, triplet, InfoNCE — every loss derived from scratch with interactive comparisons.

10 chapters12+ sims
30

Optimization & Backprop

Gradients, SGD, chain rule, computation graphs, gate patterns.

10 chapters
31

Neural Networks: Architecture

Neurons, activation functions, layers, representational power.

10 chapters
32

Neural Networks: Setup

Data preprocessing, weight init, batch norm, dropout.

10 chapters
33

Neural Networks: Training

Loss curves, learning rates, momentum, Adam, hyperparameter search.

10 chapters
34

NN Case Study: Spirals

Build a 2-layer net from scratch. Live training visualization.

10 chapters
35

Convolutional Networks

Convolution, filters, pooling, LeNet to ResNet.

10 chapters
36

Transfer Learning

Feature extraction vs fine-tuning, freezing layers.

9 chapters
37

Recurrent Networks

RNNs, LSTM, BPTT, vanishing gradients, char-level LM.

9 chapters
38

Robot Learning

Perception-action loop, model-based planning, imitation learning, diffusion policies, VLAs & foundation models.

10 chapters
39

Multi-Modal Foundation Models

CLIP, LLaVA, Flamingo, Molmo, SAM — contrastive learning, VLMs, promptable segmentation, model chaining.

10 chapters
40

3D Vision

Point clouds, meshes, SDFs, PointNet, NeRF, 3D Gaussian Splatting — from representations to neural rendering.

10 chapters
38

Regularization & Optimization

L1/L2, cross-entropy, SGD, momentum, Adam, learning rate schedules, weight initialization.

10 chapters
39

Neural Networks & Backprop

Computational graphs, chain rule, gate patterns, Jacobians — backprop from first principles.

10 chapters
40

CNNs for Classification

Convolution operation, stride, padding, pooling, receptive field, depthwise separable convolutions.

10 chapters
41

CNN Training & Architectures

Data augmentation, batch norm, dropout, AlexNet → VGG → ResNet → EfficientNet → ConvNeXt.

10 chapters
42

Recurrent Networks (2025)

Vanilla RNN, BPTT, vanishing gradients, LSTM gates, GRU, char-level LM, image captioning.

10 chapters
39

Attention & Transformers

Bahdanau attention, self-attention, multi-head, transformer blocks, positional encoding, ViT.

10 chapters
40

Detection & Segmentation

FCN, U-Net, R-CNN family, YOLO, Mask R-CNN, GradCAM, adversarial examples.

10 chapters
10

Scene Classification

Naming the whole place, not the parts: Places365, holistic context, CNN→ViT→InternImage, top-1 vs top-5.

10 chapters
10

Acoustic Scene Classification

Hearing the place from the soundscape: log-mel spectrograms, device mismatch, tiny edge models, PaSST distillation.

10 chapters
10

Conv-Transformer Mobile Hybrids

Conv locality + attention globality on a phone: MobileViT, FastViT reparameterization, EfficientViT, the latency-accuracy Pareto.

10 chapters
10

Remote-Sensing Scene Classification

Land-use from orbit: multispectral tiles, NDVI, SatMAE/SpectralGPT foundation models, and the multi-label sigmoid shift.

10 chapters
10

RGB-D Scene Classification

Indoor scenes with depth: SUN RGB-D, why geometry disambiguates layout, two-stream RGB+depth fusion.

10 chapters
10

Vision & VLM Scaling Laws

How accuracy scales with params, data, compute: power laws, ViT-G, compute-optimal, the label-noise ceiling.

10 chapters
10

Mobile & NAS Backbones

Vision on a phone: depthwise-separable conv, inverted residuals, SE, compound scaling, hardware-aware NAS.

10 chapters
10

DNN-HMM Hybrid Acoustic Models

Classic ASR: HMM sequences, GMM→DNN emissions, the posterior/prior trick, Viterbi forced alignment.

10 chapters
10

CRF & HSMM Temporal Smoothing

Kill the flicker: linear-chain CRF, Viterbi, HSMM durations, smoothing frame-level predictions into stable segments.

10 chapters
10

CNN+GNN Scene Fusion

Two-branch scenes: a holistic CNN fused with an object-relation GNN — the recipe that broke 90% on Indoor67.

10 chapters
41

Video Understanding

Two-stream networks, 3D convolutions, I3D, SlowFast, TSM, video transformers, VideoMAE.

10 chapters
42

Action Recognition

Single-frame to SlowFast to video transformers. Two-stream, 3D conv, skeleton-based, temporal detection.

10 chapters
43

Optical Flow

Brightness constancy, Lucas-Kanade, Horn-Schunck, FlowNet, RAFT. Dense motion estimation from zero.

10 chapters
44

Distributed Training

GPU hardware, data/pipeline/tensor parallelism, FSDP, ring all-reduce, 3D parallelism — training at scale.

10 chapters
43

Self-Supervised Learning

Pretext tasks, MAE, InfoNCE, SimCLR, MoCo, CPC, DINO — learning without labels.

10 chapters
44

Generative Models I

Density functions, MLE, autoregressive models, PixelCNN, autoencoders, VAEs — the full ELBO derivation.

10 chapters
45

Generative Models II

GANs, rectified flow, classifier-free guidance, latent diffusion, DiT, text-to-image & video generation.

10 chapters
System Design1

Research-backed system design lessons. Real architectures, real numbers, real tradeoffs — with interactive Canvas simulations that trace requests, visualize scale, and simulate failures.

Distributed Systems16

Practical distributed systems engineering — networking, consensus, CRDTs, transactions, caching, load balancing, resiliency patterns, observability, and deployment. Interview-grade depth with interactive simulations.

Part I

Network Foundations

TCP, TLS, flow control, congestion control, QUIC — the networking layer.

10 chapters8 sims
Part I

Service Communication

DNS, REST APIs, gRPC, idempotency — how services find and talk to each other.

10 chapters8 sims
Part II

Failure & Time

Failure detection, physical/logical/vector clocks, HLC — time in distributed systems.

10 chapters9 sims
Part II

Consensus & Replication

Raft, state machine replication, consistency models, chain replication.

10 chapters10 sims
Part II

Coordination Avoidance

CRDTs, gossip protocols, Dynamo, CALM theorem, causal consistency.

10 chapters10 sims
Part II

Distributed Transactions

ACID, isolation levels, 2PC, sagas, outbox pattern.

10 chapters10 sims
Part III

Caching & CDNs

HTTP caching, reverse proxies, CDN architecture, cache invalidation.

10 chapters10 sims
Part III

Partitioning & Storage

Range/hash partitioning, consistent hashing, blob storage.

10 chapters10 sims
Part III

Load Balancing

DNS, L4, L7 load balancing — distributing traffic across servers.

10 chapters9 sims
Part III

Data Storage & Caching

Replication, NoSQL taxonomy, caching patterns, eviction policies.

10 chapters9 sims
Part III

Service Architecture

Microservices, API gateways, service mesh, control/data planes.

10 chapters9 sims
Part III

Messaging & Events

Message queues, pub/sub, Kafka, exactly-once, backpressure.

10 chapters9 sims
Part IV

Failure Modes & Isolation

Cascading failures, redundancy, shuffle sharding, cellular architecture.

10 chapters9 sims
Part IV

Resiliency Patterns

Timeouts, retries, circuit breakers, rate limiting, load shedding.

10 chapters9 sims
Part V

Testing & Deployment

Test pyramid, chaos engineering, CI/CD, canary deploys, rollbacks.

10 chapters9 sims
Part V

Observability & Operations

Metrics, SLIs/SLOs, alerting, logs, distributed tracing, dashboards.

10 chapters9 sims
Seminal Papers2

Deep interactive lessons from foundational distributed systems papers. Every definition derived, every algorithm implemented, every proof traced.

Introduction to Algorithms32

Chapter-by-chapter deep dive into the classic algorithms textbook. Each chapter is a standalone interactive lesson with interview-grade depth, Canvas simulations, coding drills, and mastery challenges.

Chapter 2

Getting Started

Insertion sort, merge sort, algorithm analysis — the foundations of algorithmic thinking.

10 chapters10 sims
Chapter 3

Growth of Functions

Big-O, Omega, Theta — the language of algorithm efficiency.

10 chapters10 sims
Chapter 4

Divide & Conquer

Max subarray, Strassen, Master theorem — breaking problems into pieces.

10 chapters10 sims
Chapter 6

Heapsort & Priority Queues

Binary heaps, heapsort, priority queues — scheduling and selection.

10 chapters10 sims
Chapter 7

Quicksort

Partition, randomize, conquer — the fastest practical sorting algorithm.

10 chapters9 sims
Chapter 8

Sorting in Linear Time

Counting sort, radix sort, bucket sort — breaking the O(n log n) barrier.

10 chapters8 sims
Chapter 11

Hash Tables

Hashing, chaining, open addressing, perfect hashing — O(1) lookup.

10 chapters10 sims
Chapter 12

Binary Search Trees

Search, insert, delete, traverse — ordered data made fast.

10 chapters9 sims
Chapter 13

Red-Black Trees

Self-balancing BSTs — guaranteed O(log n) with coloring rules.

10 chapters9 sims
Chapter 15

Dynamic Programming

Rod cutting, LCS, edit distance, knapsack — the paradigm that dominates interviews.

11 chapters9 sims
Chapter 16

Greedy Algorithms

Activity selection, Huffman, fractional knapsack — locally optimal = globally optimal.

10 chapters9 sims
Chapter 17

Amortized Analysis

Aggregate, accounting, potential — the true cost of operation sequences.

10 chapters9 sims
Chapter 22

Graph Algorithms

BFS, DFS, topological sort, SCCs — the foundation of every graph question.

11 chapters9 sims
Chapter 23

Minimum Spanning Trees

Kruskal, Prim, Union-Find — connecting everything at minimum cost.

10 chapters9 sims
Chapter 24

Shortest Paths

Dijkstra, Bellman-Ford, DAG shortest paths — optimal routes through graphs.

10 chapters11 sims
Chapter 26

Maximum Flow

Ford-Fulkerson, max-flow min-cut, bipartite matching — optimizing network flow.

10 chapters9 sims
Chapter 9

Medians & Order Statistics

Quickselect, median of medians — finding the k-th element in linear time.

10 chapters9 sims
Chapter 18

B-Trees

Disk-optimized search trees — billions of keys with just 2-3 disk reads.

10 chapters8 sims
Chapter 19

Fibonacci Heaps

Amortized O(1) decrease-key — the theoretical champion for graph algorithms.

10 chapters9 sims
Chapter 20

van Emde Boas Trees

O(log log u) predecessor queries — when the universe is your friend.

10 chapters8 sims
Chapter 21

Disjoint Sets

Union-Find — nearly O(1) connected component queries with path compression.

10 chapters10 sims
Chapter 25

All-Pairs Shortest Paths

Floyd-Warshall, Johnson's — shortest paths between every pair of vertices.

10 chapters9 sims
Chapter 27

Multithreaded Algorithms

Fork-join, work/span, parallel merge sort — harnessing multiple cores.

10 chapters8 sims
Chapter 28

Matrix Operations

LU decomposition, least squares, Cholesky — the linear algebra engine.

10 chapters9 sims
Chapter 29

Linear Programming

Simplex, duality, LP reductions — the universal optimization framework.

10 chapters8 sims
Chapter 30

Polynomials & the FFT

DFT, FFT, convolution — O(n log n) polynomial multiplication.

10 chapters12 sims
Chapter 31

Number-Theoretic Algorithms

GCD, modular arithmetic, RSA, primality — the math behind cryptography.

10 chapters8 sims
Chapter 32

String Matching

KMP, Rabin-Karp, finite automata — finding patterns in text efficiently.

10 chapters8 sims
Chapter 33

Computational Geometry

Convex hull, closest pair, line intersection — algorithms in 2D space.

10 chapters8 sims
Chapter 34

NP-Completeness

P vs NP, reductions, SAT, TSP — the limits of efficient computation.

11 chapters11 sims
Chapter 35

Approximation Algorithms

Vertex cover, TSP, set cover — good-enough solutions with guarantees.

10 chapters9 sims
Capstone

Algorithms in Modern CS

Databases, ML, compilers, networking, crypto, graphics — where every CLRS chapter shows up.

12 chapters11 sims
Designing Data-Intensive Applications13

Chapter-by-chapter deep dive into Martin Kleppmann's DDIA. Each chapter is a standalone interactive lesson with interview-grade depth, Canvas simulations, design challenges, debug scenarios, and mastery components.

Chapter 1

Trade-Offs in Data Systems

Reliability, scalability, maintainability — the vocabulary of systems design interviews.

11 chapters9 sims
Chapter 2

Nonfunctional Requirements

SLAs, percentiles, capacity planning, load testing — quantifying system quality.

9 chapters7 sims
Chapter 3

Data Models & Query Languages

Relational, document, graph — choosing how to structure and query your data.

11 chapters10 sims
Chapter 4

Storage & Retrieval

B-trees, LSM-trees, column stores — how databases actually store and find your data.

11 chapters9 sims
Chapter 5

Encoding & Evolution

JSON, Protobuf, Avro, schema evolution — how data survives change.

11 chapters9 sims
Chapter 6

Database Replication

Leader-follower, multi-leader, leaderless — keeping copies consistent across machines.

11 chapters9 sims
Chapter 7

Data Sharding

Hash partitioning, range partitioning, rebalancing — splitting data across machines.

11 chapters9 sims
Chapter 8

Database Transactions

ACID, isolation levels, MVCC, serializability — the guarantees that keep your data correct.

11 chapters11 sims
Chapter 9

The Trouble with Distributed Systems

Network faults, clock drift, process pauses — everything that can go wrong, will.

11 chapters9 sims
Chapter 10

Consistency & Consensus

Linearizability, Raft, Paxos — how distributed nodes agree on the truth.

11 chapters11 sims
Chapter 11

Batch Processing

MapReduce, Spark, dataflow engines — processing massive datasets efficiently.

11 chapters10 sims
Chapter 12

Stream Processing

Kafka, Flink, event sourcing, windowing — processing data as it arrives.

11 chapters9 sims
Chapters 13-14

Philosophy & Ethics

Streaming philosophy, bias, privacy, responsibility — the human side of data systems.

8 chapters6 sims
Exercises & Workbooks28
WORKBOOK

CS336: Build a Language Model From Scratch

Tokenization, FLOP/memory accounting, architecture, MoE, GPUs, parallelism, scaling laws, inference, alignment — derive every number.

10 chapters51 exercises
WORKBOOK

TinyML & Efficient Deep Learning

MACs, pruning, INT8/K-means quantization, NAS, distillation, MCUNet SRAM, KV-cache, LoRA/QLoRA — squeeze models onto tiny hardware.

10 chapters50 exercises
WORKBOOK

Visual Navigation Workbook

SE(3), camera projection, features, epipolar geometry, RANSAC, Gauss-Newton, manifold retraction, VIO drift, BoW, robust SLAM.

10 chapters50 exercises
WORKBOOK

Scaling Book Workbook: Interactive Exercises

54 hands-on exercises covering scaling laws, chinchilla, compute-optimal training, loss prediction, data mixing, emergent abilities.

12 chapters54 exercises
WORKBOOK

Transformer Math Workbook

52 exercises across 5 modes: derive, trace, build, design, debug. Parameter counts, attention FLOPs, memory budgets, KV caches, throughput estimation.

10 chapters52 exercises5 exercise types
WORKBOOK

Training & Backprop Workbook

Chain rule by hand, gradient shapes, cross-entropy, Adam optimizer, batch norm, learning rate schedules, mixed precision, training diagnostics.

10 chapters54 exercises
WORKBOOK

Probability & Bayes Workbook

Bayes rule, distributions, MLE, MAP estimation, Kalman filter math, HMM forward/Viterbi, information theory, sensor fusion.

10 chapters57 exercises
WORKBOOK

RL Fundamentals Workbook

Bellman equations, value iteration, Q-learning updates, policy gradients, advantage estimation, PPO clipping, exploration strategies.

10 chapters56 exercises
WORKBOOK

Diffusion & Flow Matching Workbook

Forward process, noise schedules, DDPM loss, score functions, sampling, classifier-free guidance, latent diffusion, flow matching, ODE/SDE.

10 chapters55 exercises
WORKBOOK

Systems & Serving Workbook

Tensor/pipeline/data parallelism, continuous batching, PagedAttention, speculative decoding, quantization, cost optimization.

10 chapters58 exercises
WORKBOOK

Robotics Math Workbook

Rotation matrices, homogeneous transforms, EKF predict/update, SLAM graphs, inverse kinematics, Jacobians, PID control.

10 chapters56 exercises
WORKBOOK

Code from Scratch Workbook

Implement softmax, linear layers, attention, layer norm, positional encoding, BPE tokenizer, cross-entropy, Adam, KV cache, sampling.

10 chapters52 exercises
WORKBOOK

CLRS Algorithms Workbook

Asymptotic analysis, recurrences, sorting, hash tables, BSTs, dynamic programming, greedy, graph algorithms, shortest paths, MST.

10 chapters56 exercises
WORKBOOK

State Estimation (Advanced) Workbook

Multi-dimensional KF, EKF Jacobians, range-bearing updates, UKF sigma points, particle filters, sensor fusion, observability, noise tuning.

10 chapters56 exercises
WORKBOOK

SLAM & Navigation Workbook

EKF-SLAM, landmark observation, data association, loop closure, graph SLAM, visual odometry, RANSAC, factor graphs, occupancy grids.

10 chapters58 exercises
WORKBOOK

Graph Neural Networks Workbook

Graph basics, node embeddings, GCN/SAGE/GAT, over-smoothing, spectral theory, knowledge graphs, link prediction, graph generation.

10 chapters56 exercises
WORKBOOK

NLP Fundamentals Workbook

Word vectors, Word2Vec gradients, dependency parsing, perplexity, attention, subword tokenization, pretraining, PEFT, evaluation metrics.

10 chapters57 exercises
WORKBOOK

Signal Processing Workbook

Discrete signals, DFT by hand, quantization, Lloyd-Max, STFT, wavelets, MFCC pipeline, Bayes classifiers, SVM margins.

10 chapters60 exercises
WORKBOOK

Distributed Training Workbook

Ring AllReduce, ZeRO stages, gradient accumulation, LR scaling, pipeline parallelism, tensor parallelism, checkpointing, mixed precision.

10 chapters58 exercises
WORKBOOK

Computer Vision Workbook

Convolution math, receptive fields, pooling, anchors/IoU, NMS, ResNet, segmentation, stereo vision, homography.

10 chapters53 exercises
WORKBOOK

Linear Algebra for ML Workbook

Matrix operations, Gaussian elimination, eigenvalues, SVD, PCA, matrix calculus, least squares, norms, positive definiteness.

10 chapters59 exercises
WORKBOOK

Numerical Methods Workbook

Floating point, catastrophic cancellation, condition numbers, iterative methods, Newton's method, gradient descent, integration, sparse matrices.

10 chapters55 exercises
WORKBOOK

Information Theory Workbook

Entropy, joint/conditional entropy, KL divergence, cross-entropy loss, source coding, channel capacity, rate-distortion, VAE connection.

10 chapters59 exercises
WORKBOOK

Optimization Theory Workbook

Convexity, gradient descent, SGD variants, constrained optimization, KKT conditions, duality, proximal operators, convergence rates, Newton's method.

10 chapters57 exercises
WORKBOOK

RL Theory (CS224R) Workbook

Reward modeling, RLHF objective, PPO for RLHF, DPO loss, CQL, IQL, model-based RL, reward shaping, multi-agent RL.

10 chapters57 exercises
WORKBOOK

Deep RL Exam Prep Workbook

Mode collapse, flow matching arithmetic, REINFORCE credit assignment, actor-critic n-step returns, Q-learning TD targets, offline RL, IQL expectile loss, goal-conditioned RL.

10 chapters50 exercises
WORKBOOK

Decision Making Under Uncertainty Workbook

Bayesian networks, Dirichlet inference, value of information, Bellman backups, kernel smoothing, MCTS/UCB1, policy gradients, POMDPs, alpha vectors.

10 chapters52 exercises
WORKBOOK

Model Compression Workbook

Quantization (PTQ, QAT, STE), pruning (magnitude, structured, lottery ticket), knowledge distillation, LoRA, mixed precision, compression pipelines.

10 chapters58 exercises
Day In The Life of an Engineer15

Deep-dive lessons built around real engineering roles. Classical foundations, modern tools, real papers, interactive sims. Built for interview prep with real depth.

Role Guide

Forward Deployed Engineer

Customer discovery, rapid prototyping, production deployment at customer sites, SDK integration, debugging in customer environments, demo engineering, incident response.

13 chapters12+ sims
Role Guide

Applied AI Engineer

RAG architecture, agent design, fine-tuning, prompt engineering, evaluation, streaming, production AI infrastructure, safety & guardrails.

13 chapters13+ sims
Role Guide

Web-Scale AI & Search Engineer

Information retrieval, dense embeddings, learning to rank, recommendation systems, serving a web index, the convergence of search + recs + transformers.

13 chapters12+ sims
Role Guide

Backend & API Engineer at Scale

API design, request lifecycle, database optimization, caching, rate limiting, auth, observability, scaling patterns, developer experience.

13 chapters12+ sims
Role Guide

Infrastructure & LLM Scaling Engineer

GPU clusters, LLM serving, distributed systems, cost optimization, reliability engineering, CI/CD, monitoring — keeping AI systems running at scale.

13 chapters12+ sims
Role Guide

Developer Relations Engineer

Docs, SDKs, community, content strategy, conference talks, launches, feedback loops, measuring impact — with AI company DevRel woven throughout.

13 chapters13 sims
Role Guide

AI/GenAI Scrum Master

Experiment-driven sprints, eval-based acceptance criteria, data pipeline management, research-to-production handoffs, agentic AI project management.

13 chapters13 sims
Role Guide

Lead QE & Validation Engineer

DVA model testing, sim-to-real validation, safety compliance, CI/CD for robotics, 20 interview questions + cheat sheet.

12 chapters20 Q&A
Role Guide

3D & Multi-View Geometry Engineer

Camera models, epipolar geometry, SfM, bundle adjustment, SLAM, dense reconstruction, point tracking, DUSt3R/VGGT — the full stack for robotics perception.

13 chapters8+ sims
Role Guide

ML Performance Optimization Engineer

GPU architecture, profiling, mixed precision, distributed training, model compression, TensorRT, serving at scale, real-time autonomous driving stacks.

12 chapters11 sims
Role Guide

Robotics Simulation & Software Engineer

Physics simulation, contact models, MuJoCo/Isaac Sim, kinematics, motion planning, control, sim-to-real, RL, ROS2 — the full robotics stack.

13 chapters13 sims
Role Guide

ML Inference & Performance Engineer

Quantization, CUDA kernels, TensorRT, FlashAttention, KV-cache, distributed training, BEV perception, edge deployment, VLA — the full AV inference stack.

17 chapters16+ sims
Role Guide

Software Testing & Reliability

Test design, automation, reliability engineering, incident response, safety-critical testing, debugging, observability — onsite interview prep for robotics QE.

16 chapters16 sims
Role Guide

Agentic Engineer at Sierra

Agent SDK, RAG, evaluation, data pipelines, inference serving, monitoring, security, distributed systems — the full agentic AI platform stack.

20 chapters17+ sims
Role Guide

Frontier Lab Engineer: Day in the Life

Pre-training, RLHF, evals, safety, scaling laws, infra, on-call, paper reading, experiment tracking — what a frontier lab researcher actually does.

17 chapters16+ sims
AI Harness Engineering21

The practical toolkit for building with AI models. Embeddings, retrieval, RAG, agents, evals, safety — everything between "I have a model" and "I have a production app."

HARNESS 01

Vector Embeddings

What embeddings are, how they encode meaning, distance in high-D space, embedding models, multimodal embeddings.

10 chapters10 sims
HARNESS 02

Similarity Metrics

Cosine, dot product, Euclidean, Jaccard, learned metrics — measuring "how alike" in vector space.

9 chapters9 sims
HARNESS 03

Vector Databases

FAISS, HNSW, IVF, product quantization — storing and searching millions of vectors in milliseconds.

10 chapters10 sims
HARNESS 04

Text Chunking Strategies

Fixed, sentence, recursive, semantic, structure-aware — how to split documents for embedding and retrieval.

9 chapters9 sims
HARNESS 05

RAG

The full pipeline: chunk → embed → store → retrieve → rerank → generate. Eval, failure modes, advanced patterns.

11 chapters11 sims
HARNESS 06

Multimodal RAG

Images, tables, PDFs in RAG. Vision embeddings, ColPali, document parsing, hybrid retrieval.

9 chapters9 sims
HARNESS 07

MCP (Model Context Protocol)

USB-C for AI: tools, resources, prompts. Build MCP servers and clients. The standard for AI integrations.

10 chapters10 sims
HARNESS 08

Prompt Engineering

System prompts, few-shot, chain-of-thought, structured output, testing — systematic, not vibes-based.

10 chapters10 sims
HARNESS 09

Evaluation & Evals

LLM-as-judge, human eval, automated metrics, regression detection, A/B testing for AI systems.

10 chapters10 sims
HARNESS 10

AI Safety & Guardrails

Content filtering, jailbreak prevention, PII detection, output validation, red teaming, defense in depth.

10 chapters10 sims
HARNESS 11

Fine-Tuning in Practice

When to fine-tune vs prompt, data prep, LoRA/QLoRA, eval pipelines, deployment, cost analysis.

10 chapters10 sims
HARNESS 12

Agents & Tool Use

Function calling, ReAct, multi-step agents, state management, error recovery, guardrails for agents.

10 chapters10 sims
HARNESS 13

Agent Skills

Teach Claude a workflow in a folder of Markdown: SKILL.md, progressive disclosure, triggering descriptions, the five patterns. Build one live.

10 chapters9 sims
HARNESS 14

Agent Architectures

Single agents to multi-agent orchestration: the workflow-vs-agent divide, sequential/parallel/evaluator patterns, hierarchical vs swarm, and an interactive chooser for picking the right one.

10 chapters10 sims
HARNESS 15

Enterprise AI Transformation

The adoption playbook behind the systems you build — coalition, governance, pilot selection, metrics, scaling, and a readiness calculator. Engineer's lens, real case studies (L'Oréal, Shopify, Lotte).

10 chapters10 sims
HARNESS 16

Transcultural ML in Music

Most music AI assumes a piano. How DDSP serves Thai 7-tone tunings, Carnatic meend, and the Guqin — model vs dataset bias, and when not to ship a model at all.

9 chapters7 sims
HARNESS 17

The Founder's Playbook: AI-Native Startup

The four-stage startup lifecycle — Idea, MVP, Launch, Scale — read line by line. Interactive exit criteria, the Sean Ellis test, TAM/SAM/SOM and CAC/LTV calculators, the data-flywheel moat, and when to reach for Chat vs Cowork vs Code.

12 chapters11 sims
HARNESS 18

Loop Engineering

The fourth layer above the harness: stop prompting the agent, design the system that prompts it. The five moves of one turn, the six parts, the generator/evaluator "say no", the five failure modes, and the four silent costs.

11 chapters17 sims
HARNESS 19

Harness Engineering

The system around the model — three design patterns, a coding agent's tool belt, and why the deployment layer rivals raw intelligence.

10 chapters9 sims
HARNESS 20

Harness Optimization

From prompts to optimizer code: ACE playbooks, MCE bi-level skill evolution, Meta-Harness Pareto search, ADAS, and AFlow's MCTS.

10 chapters10 sims
HARNESS 21

Self-Improving Harnesses

STOP's improver improving itself, AlphaEvolve, the Darwin Gödel Machine, SIA — and seven bottlenecks on the road to RSI.

11 chapters11 sims
Teaching Sessions1

Record yourself teaching on a whiteboard with optional camera. Export as MP4. The ultimate Feynman test: if you can teach it, you understand it.