Index Ai Academy
Foundation · Probability

Counting: choices, orders and groups

The counting principle, factorials, ordered picks and groups, and why attention's cost grows with the square of the length.

About 43 minutes · Free lesson

What it covers6

Why it matters

Probability starts with counting: the chance of something is often the number of ways it can happen divided by the number of ways anything can happen. Listing every possibility works for three or four, and fails long before a hundred. A handful of counting rules give the number at once: multiply the choices, use a factorial for orderings, divide out repeats for groups. They also explain some hard facts about AI: why a language model can never consider every possible sentence, and why attention, which compares every word with every other, gets expensive as text gets longer. This lesson builds the rules from slots, checks them in Python, and applies them to those two questions.

What you'll be able to do

  • Count possibilities by filling slots left to right and multiplying the choices
  • Count orderings with factorials, ordered picks with n!/(n − r)!, and groups with n!/(r!(n − r)!)
  • Count with math.factorial, math.perm and math.comb, and confirm small cases with a loop
  • See why a model can't try everything: sentences grow like vⁿ, and attention's table of scores like n², in Python and PyTorch
Quick check
Have a guess before you start. A camp menu has 3 breakfasts and 4 dinners. How many different breakfast-and-dinner days are there?
  1. 34
  2. 12
  3. 7
  4. 3
Show the answer

12

Each of the 3 breakfasts can go with each of the 4 dinners: 3 × 4 = 12.

The math: fill the slots, multiply

Purpose. Know how many possibilities there are without listing them.

The counting principle. When a choice is made in steps, the number of outcomes is the product of the number of choices at each step. Fill the slots left to right:

ways=n1×n2×⋯×nk\text{ways} = n_1 \times n_2 \times \cdots \times n_k

In plain words: for each way of filling the first slot, there are n₂ ways to fill the second, and so on.

A 4-digit PIN has 10 choices per slot: 10 × 10 × 10 × 10 = 10⁴ = 10,000. A code of 3 letters then 2 digits has 26 × 26 × 26 × 10 × 10 = 1,757,600.

Quick check
From lesson 10: what is 2¹⁰?
  1. 20
  2. 1,024
  3. 100
  4. 512
Show the answer

1,024

2 × 2 × … ten times = 1,024. It's also the number of yes-or-no answer sheets for 10 questions: two choices per slot.

When choices shrink. Choosing a gold, silver and bronze winner from 5 hikers, A to E: 5 choices for gold, then 4 left for silver, then 3 for bronze.

The picture. Step through the slots two ways at once: as boxes of numbers, and as a tree of choices that grows left to right, one column per slot. A ? is a slot still to fill. At each step the top branch splits once for every hiker who could take the next place. The whole tree would have 60 ends, too many to draw, so a box such as +3 stands for 3 more choices that branch the same way.

Try it now

Then set the choices for each place to 5, 4, 3, 2, 1 and press Use these numbers: every hiker in a line, 5! = 120 orders, and the tree grows two more columns.

Symbols and formula. Three counts come up again and again:

n!=n×(n−1)×⋯×1P(n,r)=n!(n−r)!C(n,r)=n!r! (n−r)!n! = n \times (n-1) \times \cdots \times 1 \qquad P(n, r) = \frac{n!}{(n-r)!} \qquad C(n, r) = \frac{n!}{r!\,(n-r)!}

In plain words: n! ("n factorial") orders all n things; P(n, r) picks r of them in order; C(n, r) ("n choose r") picks r of them as a group, ignoring order.

5 hikers in a line: 5! = 120. Gold, silver, bronze: P(5, 3) = 60. A team of 3: C(5, 3) = 10. Each group of 3 can be ordered in 3! = 6 ways, so the ordered count, 60, counts every group 6 times: 60 ÷ 6 = 10. Dividing out the repeats is where the r! comes from: one pair in two orders is still one pair.

Quick check
How many ways can 4 tent spots be assigned to 4 groups, one each?
  1. 4
  2. 24
  3. 16
  4. 256
Show the answer

24

4 choices for the first spot, 3 for the next, then 2, then 1: 4! = 24.

Quick check
A committee of 2 is picked from 6 people. How many committees are possible?
  1. 12
  2. 30
  3. 36
  4. 15
Show the answer

15

Ordered picks: 6 × 5 = 30. Each committee was counted twice (A then B, B then A): 30 ÷ 2 = 15.

In your own words
In your own words: why is a group of r counted r! times among the ordered picks?

Write a sentence or two, then compare it with a good answer in the lesson.

In Python: count, then confirm by listing

Purpose. Compute counts with one call, and check small ones by listing every case.

import math

print(math.factorial(5), math.perm(5, 3), math.comb(5, 3))
print(10 ** 4, 26 ** 3 * 10 ** 2)
It runs right here in your browser, with nothing to install.

math.factorial(n) is n!, math.perm(n, r) is ordered picks, and math.comb(n, r) is groups. For small cases, a loop that lists every possibility is the best check that the right formula was chosen.

Code flow. Here is that check for 3 items, numbered 0, 1 and 2, taken two at a time. First the formulas count, then two loops list. Step through and watch the two lists fill: a pair goes into ordered whenever its two items differ (i ≠ j), but into groups only in one order (i < j).

The loops list 6 ordered pairs and 3 groups, the same two counts the formulas printed first. Each group shows up twice in ordered, once in each order. The same check for 4 items, counting instead of listing:

items = ["tent", "tarp", "stove", "map"]
ordered = 0
groups = 0
for i in range(len(items)):
    for j in range(len(items)):
        if i != j:
            ordered = ordered + 1
        if i < j:
            groups = groups + 1
print(ordered, groups)

12 ordered pairs and 6 groups: math.perm(4, 2) and math.comb(4, 2) agree.

A named mistake: perm for comb. "How many teams" means order doesn't matter: math.comb. "How many ways to pick a captain and a vice-captain" means it does: math.perm. Asking "would swapping two picks give a different answer?" decides which. Using perm for a team overcounts by r!.

Worked example

Worked example: count and verify four cases

Compute four counts with the formulas, and confirm the smallest by listing.

import math

pins = 10 ** 4
lineups = math.factorial(6)
podiums = math.perm(8, 3)
teams = math.comb(8, 3)
print(pins, lineups, podiums, teams, podiums // teams)

count = 0
for a in range(5):
    for b in range(a + 1, 5):
        for c in range(b + 1, 5):
            count = count + 1
print(count, math.comb(5, 3))
In the lesson you can change this and run it.

10,000 PINs, 720 line-ups of 6, 336 podiums from 8 and 56 teams of 3 from 8. podiums ÷ teams = 6 = 3!: each team is counted 6 times among the podiums. The three nested loops list every group of 3 from 5 in increasing order, a < b < c, so each group appears exactly once: 10, matching math.comb(5, 3).

Quick check
What does this print?
import math

print(math.perm(6, 2), math.comb(6, 2), math.factorial(4))
  1. 36 15 24
  2. 15 30 24
  3. 30 15 10
  4. 30 15 24
Show the answer

30 15 24

Ordered pairs from 6: 6 × 5 = 30. Groups of 2: 30 ÷ 2 = 15. 4! = 24.

Drill

Pairs, two ways

Fix the loop so pairs counts each unordered pair of the 6 items exactly once. Then set formula to the same count from math.comb, and ordered to the ordered count from math.perm.

import math

pairs = 0
for i in range(6):
    for j in range(6):
        pairs = pairs + 1
formula = 0
ordered = 0
print(pairs, formula, ordered)
In the lesson, your code is checked as you write it.
In your own words
How do you decide between math.perm and math.comb for a counting question?

Write a sentence or two, then compare it with a good answer in the lesson.

In AI and ML: why a model can't try everything

Purpose. Counting shows where brute force runs out.

Sentences. A language model with a vocabulary of v words could write vⁿ different sentences of n words: n slots, v choices for each. Here is one entry from the trail club's camp log. .split() cuts it into words, and set(log) keeps one copy of each word (lesson 28): the log's vocabulary.

log = "the fire was warm and the tent was dry so we all slept well".split()
v = len(set(log))
print(len(log), v)
for n in [1, 2, 3, 4, 5]:
    print(n, v ** n)

The entry has 14 words but only 12 different ones. From those 12 words you could write 12 one-word sentences, 144 of two words, 1,728 of three and 248,832 of five.

The picture. The same counts as bars, first for 1 to 3 words, then for 1 to 5 on one scale. Keep your eye on the 3-word bar when you press Next.

Ten-word sentences from the same 12 words number about 62 billion (12¹⁰). Real vocabularies have 50,000 or more words, so a model can never check every sentence; it builds one word at a time instead (lesson 91).

sentences=vnpairs of words=n(n−1)2\text{sentences} = v^n \qquad \text{pairs of words} = \frac{n(n-1)}{2}

In plain words: sentences grow exponentially with length; pairs grow with the square of the length.

Attention's cost. A transformer's attention compares every word with every other word in its input. For n words that's n(n − 1)/2 pairs: 45 for 10 words, 4,950 for 100, and 499,500 for 1,000. Ten times the text means about a hundred times the comparisons.

Quick check
What does this print?
pairs = []
for n in [10, 100, 1000]:
    pairs.append(n * (n - 1) // 2)
print(pairs)
  1. [45, 4950, 499500]
  2. [90, 9900, 999000]
  3. [45, 450, 4500]
  4. [100, 10000, 1000000]
Show the answer

[45, 4950, 499500]

n(n − 1)/2 pairs: each word paired with every other, counted once. Ten times the words gives about a hundred times the pairs.

In PyTorch. A real attention layer keeps its comparisons in a table of scores, a row for each word and a column for each word it looks at: T × T numbers for a text of T words. torch.ones(T, T) makes a table that size (every entry 1, standing in for a score), and .numel() counts its entries. Here it is for the camp-log entry:

import torch

log = "the fire was warm and the tent was dry so we all slept well".split()
T = len(log)
pairs = T * (T - 1) // 2
scores = torch.ones(T, T)
print(T, pairs, scores.shape, scores.numel())
print(2 * pairs + T)

The table holds 196 numbers for 14 words, and the plain-Python line under it gives the same 196: every pair in both orders plus each word with itself, 2 × 91 + 14. So the table grows like the pairs, with the square of the length. The math part's flow has a PyTorch tab built the same way: torch.ones([5, 4, 3]) (a list of sizes works as well as separate numbers) has one axis per slot and one entry per way, and .numel() multiplies 5 × 4 × 3 = 60.

Why this matters in practice. Doubling a model's context length roughly quadruples the work attention does, which is why long-context models need special tricks to stay affordable. And since sentences can't be listed, generating text is always a matter of choosing well, step by step, among the next word's options. With a vocabulary of 50,000 words, a sentence of 10 words has 50,000¹⁰ possible versions, about 10⁴⁷: a 1 followed by 47 zeros. Even a computer scoring a billion sentences a second would need far longer than the age of the universe to score them all. So no model tries. Search methods such as beam search keep a few of the best partial sentences, say 4, so each step scores 4 × 50,000 = 200,000 options rather than every complete sentence. Counting gives the size of a search before you write any code, and tells you whether "try them all" is a plan or a fantasy.

Drill

Sentences and pairs

For a vocabulary of 12 words, set sentences to the number of sentences of lengths 3, 5 and 10, as a list. Then set pairs to the number of word pairs in texts of 10, 100 and 1,000 words, as a list, and ratio to pairs for 1,000 words divided by pairs for 100, rounded to 1 place.

sentences = []
pairs = []
ratio = 0
print(sentences, pairs, ratio)
In the lesson, your code is checked as you write it.
In your own words
Why does a language model build a sentence one word at a time instead of choosing among all possible sentences?

Write a sentence or two, then compare it with a good answer in the lesson.

Goes deeper in

Practice what you just read

  • 5 code examples you can edit and run
  • a practice exercise that checks your code
  • a 5-question quiz
  • 9 review cards for spaced repetition

Try a quiz question

How many ways can a president and a treasurer be chosen from 8 people?

  1. 56
  2. 28
  3. 64
  4. 16

Answer it at the end of the lesson to see if you're right, and why.

Before this lesson

Keep learning

Where this lesson leads.

575 lessons in thirteen phases. 176 are free without an account, and a free account opens the rest. See every lesson.

Foundation · about 96 hours of study

Foundation math

From whole numbers to the math inside a neural network: each idea by hand, in Python, and in AI 118 lessons All free
  1. Numbers and place value FreeWhat a digit's place is worth, the number line, and Python as a calculator that never tires.36 min
  2. Multiplying and dividing: groups FreeEqual groups, quotient and remainder, and cutting training data into batches.35 min
  3. Order of operations, and names for numbers FreeBrackets, powers, then multiply and divide, then add and subtract; variables hold numbers.36 min
  4. Numbers below zero FreeThe number line through zero, sign rules, distance from zero, and an error's direction and size.37 min
  5. Fractions: equal parts, equal fractions, adding FreePieces of a whole, equal fractions, and re-cutting to add; accuracy as a fraction.45 min
  6. Multiplying and dividing fractions FreeA share of a share, dividing by flipping, and why chances in a row get small.35 min
  7. Decimals and rounding FreeFractions in place value, rounding, and why 0.1 + 0.2 isn't exactly 0.3.36 min
  8. Percentages FreePer hundred, percent change, percentage points, and the accuracy and error rates of models.35 min
  9. Ratios, rates and proportion FreeSharing in a ratio, rates per unit, scaling, and splitting data 80/10/10.35 min
  10. Powers and roots FreeRepeated multiplying, the laws of exponents, square roots, and squared error.43 min
  11. Very big and very small numbers FreePowers of ten, e-notation like 3e-4, orders of magnitude, and a model's size in memory.36 min
  12. Bits, bytes and floats: how numbers are stored FreeBase 2, what n bits can hold, a float as sign, power and digits, and a tensor's memory.35 min
  13. Review: find the slip (numbers) FreeBroken calculations to fix, mixed practice from the whole module, and its formula sheet.34 min
  14. Letters for numbers FreeVariables, terms and like terms, substitution, and a model's formula as a Python function.41 min
  15. Expanding and factoring FreeBrackets as areas, pulling out common factors, and a squared error expanded into a bowl.41 min
  16. Solving equations: keep the balance FreeThe same move on both sides, undoing steps in reverse, and running a model's formula backward.41 min
  17. Rearranging a formula FreeMaking any letter the subject, round-trip checks, and planning a training run.36 min
  18. Inequalities and true-or-false FreeRanges, flipping the sign, if and elif, and the thresholds, clips and ReLU inside a model.42 min
  19. Functions: a rule from input to output Freef(x) notation, tables and graphs, domain and range, loops over lists, and loss as a function.42 min
  20. Chaining and undoing functions Freeg(f(x)), inverses that undo, scaling and unscaling, and a network as a chain.42 min
  21. Straight lines: slope and intercept FreeRise over run, the line through two points, and a model's weight and bias as slope and intercept.42 min
  22. Two equations, two unknowns FreeSubstitution and elimination, where lines cross, and why models are fitted, not solved.42 min
  23. Quadratics: bowls and their lowest point FreeRoots by factoring and by formula, the discriminant, the vertex, and the best weight of a model.42 min
  24. Exponentials: growth and decay FreeThe same factor each step, the number e, and decaying learning rates.43 min
  25. Logarithms: the power question FreeTo what power?, the laws of logs, and adding log-probabilities instead of multiplying.45 min
  26. Sequences and Σ FreeArithmetic and geometric sequences, Σ as a loop that adds, and mean squared error.40 min
  27. Weighted and moving averages FreeWeights that add to 1, a window average, the exponential moving average, and smoothing a loss curve.35 min
  28. Sets, and reading the symbols FreeSets, union and intersection, ℝⁿ and f : ℝⁿ → ℝᵐ, and reading a paper's first equation.41 min
  29. Review: find the slip (algebra) FreeBroken algebra and code to fix, mixed practice from the whole module, and its formula sheet.34 min
  30. Transposing step by step: undo in reverse order FreeA formula's layers, the opposite of each, peeling from the outside in, conditions, and round-trip checks.35 min
  31. Formulas with fractions: clear them first FreeClearing denominators, cross-multiplying, a letter in the bottom, and reciprocal formulas like 1/R = 1/R₁ + 1/R₂.35 min
  32. When the letter appears twice: collect and factor FreeGathering the letter on one side, factoring it out, dividing by the bracket, and rearranging precision and F1.35 min
  33. Squares, roots and powers FreeSquare roots with ±, squaring to undo a root, cube and nth roots, and the √d inside attention.40 min
  34. When the letter is squared and plain: quadratics FreeRearranging into ax² + bx + c = 0, the formula, completing the square, and choosing the root that fits.35 min
  35. Letters in the power: exponentials and logarithms FreeTaking logs to bring a power down, exponentiating to free a letter, and the logit that undoes the sigmoid.40 min
  36. Rearranging inequalities: when the sign flips FreeThe four moves on an inequality, negatives and reciprocals that flip it, logs below 1, and sample sizes.35 min
  37. Inverse functions, lines and scaling FreeUndoing a function by making x the subject, domains that allow it, line forms, and unscaling predictions.40 min
  38. Sequences and sums: find n, a, d and r FreeEvery letter of the arithmetic and geometric formulas as the subject, and an average's window from its decay.40 min
  39. Combining formulas: substitute and eliminate FreePutting one formula inside another to remove a letter, and sizing a training run from a time or money budget.35 min
  40. Review: rearrange anything (algebra) FreeBroken rearrangements to fix, mixed practice from the whole algebra block, a strategy chart, and its formula sheet.35 min
  41. Coordinates and distance FreePoints as (x, y), Pythagoras, distance in any number of coordinates, and nearest neighbors.42 min
  42. Angles, degrees and radians FreeMeasuring a turn, converting units, arcs, and the angle between two arrows.36 min
  43. Sine and cosine FreeAcross and up: right triangles, the unit circle, and directions and rotations.42 min
  44. Waves: height, speed and shift FreeAmplitude, period and phase, sampling a wave, and position codes in transformers.37 min
  45. Review: find the slip (shapes and waves) FreeBroken geometry and code to fix, mixed practice from the module, and its formula sheet.29 min
  46. Angles, triangles and waves: rearranging with sin, cos and tan FreeSides and angles from SOH CAH TOA, principal values, arcs, polar coordinates, and a wave's time and period.45 min
  47. Vectors: lists of numbers that point FreeA list, a point and an arrow; components, dimension, NumPy arrays, and data as vectors.42 min
  48. Adding and scaling vectors FreeComponent-wise sums, scalar multiples, mixes, and the weight update w − ηg.43 min
  49. Length and distance: norms FreeL2, L1 and max norms, unit vectors, distance, and the norms inside losses and clipping.42 min
  50. The dot product FreeMultiply in pairs and add; a·a as squared length; a @ b; one neuron as a dot product.43 min
  51. Angle, cosine similarity and projection FreeDirection without length, perpendicular vectors, shadows, and search by meaning.45 min
  52. Perpendicular directions: bases and coordinates FreeOrthonormal directions, coordinates as dot products, rebuilding a vector, and taking a direction out.35 min
  53. Matrices: tables of numbers FreeRows, columns and entries; two-dimensional arrays and slices; a data set as a matrix.41 min
  54. Adding, scaling and transposing matrices FreeEntry-by-entry sums and scalings, the transpose, mean images, and tables turned around.36 min
  55. A matrix times a vector FreeEach row dotted with the vector, the size rule, A @ x, and a layer acting on one example.42 min
  56. Matrix multiplication FreeRow meets column, the size rule, order matters, and a batch through a layer: X W + b.45 min
  57. Identity, diagonal, triangular and symmetric FreeFour patterns and their jobs: rescaling, distance tables, and the causal mask.37 min
  58. Matrices as movements of space FreeStretches, flips, shears and rotations; columns as landing spots; chaining; a layer moving data.42 min
  59. Determinant and inverse FreeThe area factor ad − bc, singular matrices, undoing a matrix, and redundant features.42 min
  60. Solving systems by elimination FreeAugmented matrices, row moves, back-substitution, the three outcomes, and fitting weights.45 min
  61. Independence, span and rank FreeRedundant vectors, span, rank, outer products, and low-rank updates like LoRA.42 min
  62. Eigenvectors: directions that only stretch FreeAv = λv, checking candidates, np.linalg.eig, and the main direction of a data cloud.41 min
  63. SVD and PCA: the important directions first FreeA matrix as pieces by size, the best rank-k copy, compression, and PCA maps of data.45 min
  64. Tensors, shapes and broadcasting FreeAxes and shapes, reshape, axis=, broadcasting, and shapes through a small model.43 min
  65. One-hot vectors and lookup tables: embeddings FreeA one-hot row picks a table's row, a sentence looked up as vectors, and a score for every word.35 min
  66. Sliding dot products: convolution FreeA small pattern slid along a list and over a grid, output sizes, stride and padding, and shared weights.40 min
  67. Review: find the slip (vectors and matrices) FreeShape slips to fix, mixed practice from the module, and its formula sheet with shapes.36 min
  68. Vector equations: rearranging with arrows FreeSolving for a vector or a scale, recovering a gradient from two weight snapshots, and an angle from a dot product.40 min
  69. Matrix equations: multiply by the inverse, on the correct side FreeNo dividing by a matrix: inverses from the left or right, order, transposes, and the normal equations.40 min
  70. Rates of change FreeChange over change, secants, rates from readings with np.diff, and reading a loss curve.41 min
  71. Limits: what a value heads toward FreeApproaching from both sides, holes, corners and jumps, epsilon, and judging activations.37 min
  72. The derivative: slope at a point FreeShrinking secants, f′(x) and dy/dx, Δy ≈ f′Δx, and which way a weight should move.45 min
  73. Derivative rules FreeConstants, powers, multiples, sums, eˣ and ln x, checked numerically, and a squared error's slope.41 min
  74. The product rule and the chain rule FreeProducts and chains of functions, forward then backward, and one neuron's dL/dw.45 min
  75. Lowest points and curvature Freef′ = 0, the second-derivative test, tables, local and global minima, and convex losses.43 min
  76. Standing in for a curve near a point FreeTangent lines, a curvature term, growing error, and why learning rates must be small.37 min
  77. Partial derivatives FreeOne input at a time, ∂, nudging in Python, and the two slopes of a line's loss.41 min
  78. The gradient FreeAll the slopes as a vector, steepest climb and descent, a numerical gradient, and the loss gradient.42 min
  79. Gradient descent FreeStepping against the gradient, the learning rate, a training loop, and fitting a line.45 min
  80. Faster descent: momentum and Adam FreeMomentum and Adam by hand, the bias correction, and three optimizers down one narrow valley.40 min
  81. The chain rule with many paths: backpropagation by hand FreeGraphs, multiply along paths and add across them, gradient checks, and a 2-2-1 network by hand.45 min
  82. Automatic differentiation: how PyTorch finds every slope FreeA tape of steps walked backward, and requires_grad, backward(), .grad, zero_grad() and no_grad().40 min
  83. Jacobians and Hessians FreeSlopes of vector functions, a layer's Jacobian, second partials, and bowls, domes and saddles.42 min
  84. Integrals: adding up thin slices FreeStrip and trapezoid sums, summing undoes differencing, probability as area, and AUC.43 min
  85. Review: find the slip (calculus) FreeCalculus slips to fix, mixed practice from the module, and its formula sheet.36 min
  86. Rearranging in calculus: slopes, steps and constants FreeSetting a slope to zero, Newton's step, the chain rule turned around, constants from a point, and learning rates.45 min
  87. Counting: choices, orders and groups FreeThe counting principle, factorials, ordered picks and groups, and why attention's cost grows with the square of the length.43 min
  88. Probability: from counts to chances FreeEqually likely outcomes, the complement, distributions, seeded simulation, and a next-word model.43 min
  89. Combining events: and, or, independent FreeThe or rule, independence, probability trees, drawing with and without replacement, and chains of steps.42 min
  90. Conditional probability and Bayes' rule FreeConditionals from tables, why P(A | B) isn't P(B | A), Bayes' rule, and precision and recall.45 min
  91. Random variables and distributions FreeUncertain numbers, the probability of a range, sampling by slices, and sampling versus argmax.42 min
  92. Expected value, variance and standard deviation FreeProbability-weighted averages, spread in the data's own units, and a training loss as an expected value.42 min
  93. Bernoulli, binomial, uniform and normal FreeYes-or-no trials, counts of yeses, the 68-95-99.7 rule, z-scores, and an accuracy's error bar.45 min
  94. Two variables at once: covariance and correlation FreeJoint tables and marginals, covariance, correlation, the covariance matrix, and features that carry the same information.45 min
  95. Review: find the slip (probability) FreeProbability slips to fix, mixed practice from the module, and its formula sheet.36 min
  96. Probability formulas: solve for the missing chance FreeConditionals, Bayes' rule for the prior, the or rule, expected values, variance, and z-scores turned around.45 min
  97. Describing data: center, spread and scale FreeMean, median and mode, variance and standard deviation, percentiles, outliers, and standardizing features.42 min
  98. Samples, the law of large numbers and the central limit theorem FreeSamples and populations, the standard error σ/√n, the bell of averages, and error bars on test scores.43 min
  99. Is the difference real? Comparing two models FreePaired differences, a bootstrap interval, what a p-value is and isn't, and why many tries make lucky wins.40 min
  100. Maximum likelihood FreeLikelihoods as products, log-likelihoods as sums, grid search, underflow, and the negative log-likelihood loss.45 min
  101. Linear regression: the best line FreeResiduals, least squares by formula, three roads to one line, the matrix form, and train versus test error.45 min
  102. Logistic regression: a line that answers yes or no FreeThe sigmoid, the log loss, its slope p − y, a training loop, and choosing a threshold.45 min
  103. Scoring a classifier: confusion table, precision, recall, F1 FreeThe four cells, accuracy, precision, recall and F1, and what moving a threshold does to each.35 min
  104. Softmax: scores into probabilities FreeExponentiate and share out, shifting and temperature, the safe softmax, and a classifier's last step.43 min
  105. Surprise, entropy and cross-entropy FreeSurprise in bits, entropy, cross-entropy and KL divergence, perplexity, and a language model's loss.45 min
  106. KL divergence: how far one distribution is from another FreeKL by hand, cross-entropy as entropy plus KL, why it isn't symmetric, and a soft-target loss.35 min
  107. Review: find the slip (statistics and information) FreeStatistics slips to fix, mixed practice from the module, and its formula sheet.38 min
  108. Statistics and information: n, z, logits and bits FreeSample sizes from error bars, inputs from probabilities, logit gaps from softmax, and chances from bits and losses.40 min
  109. One neuron, start to finish FreeWeighted sum, activation, loss, slopes and update, by hand and in code, and what one straight line can and can't split.45 min
  110. A layer and a small network as matrices FreeTwo layers on a batch, the shape of every matrix, adding a hidden unit, and XOR solved.45 min
  111. Training a small network FreeThe backward pass in matrix form, gradient checks, a full training loop, and reading loss curves.45 min
  112. The same network in PyTorch Freenn.Sequential with the NumPy weights loaded, the same loss and slopes, then SGD and mini-batches.45 min
  113. Attention as matrix math FreeScaled dot-product scores, the causal mask, a row-by-row softmax, and blending value vectors.45 min
  114. Residual connections and layer norm FreeAdding the input back and why its slope keeps learning alive, layer norm by hand, and add and norm.40 min
  115. A transformer block, start to finish FreeA sentence through embeddings, attention, add and norm, a small network and next-word scores, every shape named.45 min
  116. Planning a model: rearranging the formulas of a working network FreeLayer widths from parameter counts, sizes from memory and compute, the attention scale, and a training plan.40 min
  117. Final test: transpose any formula FreeSlips to fix from every topic, mixed practice of every kind of rearranging, the strategy chart and the full sheet.45 min
  118. Capstone: a sign reader from the formulas FreeBuild, train and test a trail-sign classifier with NumPy alone, and report its accuracy with an error bar.45 min
Part of a free course 118 lessons

The Foundation Math For Ai & ML: from arithmetic to attention

Learn the math behind AI from arithmetic up: numbers, algebra, vectors and matrices, calculus, probability and statistics, each shown as numbers moving and checked in Python, ending with a network you build and train from the formulas.

Never trained an AI? Teach a small one to write, right in your browser. It takes about 20 seconds and needs no account. Teach it to write