The network after 766 updates
Train it
0.001566total error E
0.0506, 0.9515outputs (want 0.01, 0.99)
1. Forward pass: compute the answer
net_h1 = w1·i1 + w2·i2 + b1 = 0.2707·0.05 + 0.4414·0.10 + 0.35 = 0.407674
h1 = σ(net_h1) = 0.600530 h2 = σ(0.422305) = 0.604035
net_o1 = w5·h1 + w6·h2 + b2 = −2.931391 o1 = 0.050623
net_o2 = w7·h1 + w8·h2 + b2 = 2.976555 o2 = 0.951504
E = ½(t1 − o1)² + ½(t2 − o2)² = 0.000825 + 0.000741 = 0.001566117
What to notice
Output weights are easy. w5's blame is just three numbers multiplied: how wrong o1 is, how steep o1's sigmoid is, and how big h1 was (a big input means the weight mattered more).
Hidden weights reuse that work. w1 can't see the error directly. Its blame arrives through both outputs: output 1's blame times w5, plus output 2's blame times w7. Then the same local slopes. That reuse is what makes backprop fast enough for billions of weights.
The check column agrees. "Nudge and measure" gives the same numbers without any calculus, but it needs two extra forward passes per weight. For an LLM that would be billions of passes per step; backprop needs one.
2. Backward pass: blame for every weight
| Weight | Local slopes multiplied together (chain rule) | ∂E/∂w | Check (nudge) | New value = w − 0.50 × ∂E/∂w |
|---|---|---|---|---|
| w5 | (o₁ − t₁) 0.04062 × o₁(1 − o₁) 0.04806 × h₁ 0.60053 | 0.0011725 | 0.0011725 | −2.94674 → −2.94732984 |
| w6 | (o₁ − t₁) 0.04062 × o₁(1 − o₁) 0.04806 × h₂ 0.60403 | 0.0011793 | 0.0011793 | −2.91669 → −2.91728168 |
| w7 | (o₂ − t₂) −0.03850 × o₂(1 − o₂) 0.04614 × h₁ 0.60053 | −0.0010668 | −0.0010668 | 1.94360 → 1.94413103 |
| w8 | (o₂ − t₂) −0.03850 × o₂(1 − o₂) 0.04614 × h₂ 0.60403 | −0.0010730 | −0.0010730 | 2.00215 → 2.00268419 |
| w1 | δo₁·w5 + δo₂·w7 −0.00921 × h₁(1 − h₁) 0.23989 × i₁ 0.05000 | −0.0001104 | −0.0001104 | 0.27069 → 0.27074976 |
| w2 | δo₁·w5 + δo₂·w7 −0.00921 × h₁(1 − h₁) 0.23989 × i₂ 0.10000 | −0.0002208 | −0.0002208 | 0.44139 → 0.44149953 |
| w3 | δo₁·w6 + δo₂·w8 −0.00925 × h₂(1 − h₂) 0.23918 × i₁ 0.05000 | −0.0001106 | −0.0001106 | 0.36922 → 0.36927380 |
| w4 | δo₁·w6 + δo₂·w8 −0.00925 × h₂(1 − h₂) 0.23918 × i₂ 0.10000 | −0.0002213 | −0.0002213 | 0.53844 → 0.53854761 |
At the start with learning rate 0.5 you should see w5 → 0.35891648 and w1 → 0.14978072, the same as the original worked example. δo means (o − t)·o(1 − o), the output neuron's blame.
The code that just ran
Everything on this page came from calling this code, which lives at the top of this same file.
Show or hide the source
final class TinyNet
{
public array $w = ['w1' => 0.15, 'w2' => 0.20, 'w3' => 0.25, 'w4' => 0.30,
'w5' => 0.40, 'w6' => 0.45, 'w7' => 0.50, 'w8' => 0.55,
'b1' => 0.35, 'b2' => 0.60]; // biases stay fixed, as in the original example
public array $in = [0.05, 0.10];
public array $target = [0.01, 0.99];
public function forward(): array
{
$w = $this->w; [$i1, $i2] = $this->in; [$t1, $t2] = $this->target;
$nh1 = $w['w1'] * $i1 + $w['w2'] * $i2 + $w['b1']; $h1 = sigmoid($nh1);
$nh2 = $w['w3'] * $i1 + $w['w4'] * $i2 + $w['b1']; $h2 = sigmoid($nh2);
$no1 = $w['w5'] * $h1 + $w['w6'] * $h2 + $w['b2']; $o1 = sigmoid($no1);
$no2 = $w['w7'] * $h1 + $w['w8'] * $h2 + $w['b2']; $o2 = sigmoid($no2);
$E1 = 0.5 * ($t1 - $o1) ** 2;
$E2 = 0.5 * ($t2 - $o2) ** 2;
return compact('i1', 'i2', 't1', 't2', 'nh1', 'h1', 'nh2', 'h2', 'no1', 'o1', 'no2', 'o2', 'E1', 'E2') + ['E' => $E1 + $E2];
}
public function loss(): float { return $this->forward()['E']; }
// BACKPROPAGATION. Each gradient is a product of simple local slopes:
// (o − t) how the error changes with the output
// o·(1 − o) how the sigmoid output changes with its input
// h or i how that input changes with the weight
// Hidden weights get the SUM of the blame flowing back from both outputs.
public function gradients(): array
{
$f = $this->forward(); $w = $this->w;
$do1 = ($f['o1'] - $f['t1']) * $f['o1'] * (1 - $f['o1']); // δ of output 1
$do2 = ($f['o2'] - $f['t2']) * $f['o2'] * (1 - $f['o2']); // δ of output 2
$back1 = $do1 * $w['w5'] + $do2 * $w['w7']; // blame reaching h1
$back2 = $do1 * $w['w6'] + $do2 * $w['w8']; // blame reaching h2
$s1 = $f['h1'] * (1 - $f['h1']);
$s2 = $f['h2'] * (1 - $f['h2']);
$g = fn($factors) => ['factors' => $factors, 'value' => array_product(array_column($factors, 1))];
return [
'w5' => $g([['(o₁ − t₁)', $f['o1'] - $f['t1']], ['o₁(1 − o₁)', $f['o1'] * (1 - $f['o1'])], ['h₁', $f['h1']]]),
'w6' => $g([['(o₁ − t₁)', $f['o1'] - $f['t1']], ['o₁(1 − o₁)', $f['o1'] * (1 - $f['o1'])], ['h₂', $f['h2']]]),
'w7' => $g([['(o₂ − t₂)', $f['o2'] - $f['t2']], ['o₂(1 − o₂)', $f['o2'] * (1 - $f['o2'])], ['h₁', $f['h1']]]),
'w8' => $g([['(o₂ − t₂)', $f['o2'] - $f['t2']], ['o₂(1 − o₂)', $f['o2'] * (1 - $f['o2'])], ['h₂', $f['h2']]]),
'w1' => $g([['δo₁·w5 + δo₂·w7', $back1], ['h₁(1 − h₁)', $s1], ['i₁', $f['i1']]]),
'w2' => $g([['δo₁·w5 + δo₂·w7', $back1], ['h₁(1 − h₁)', $s1], ['i₂', $f['i2']]]),
'w3' => $g([['δo₁·w6 + δo₂·w8', $back2], ['h₂(1 − h₂)', $s2], ['i₁', $f['i1']]]),
'w4' => $g([['δo₁·w6 + δo₂·w8', $back2], ['h₂(1 − h₂)', $s2], ['i₂', $f['i2']]]),
];
}
// Sanity check without any calculus: nudge the weight up and down a hair
// and see how much the loss moves. Should match backprop almost exactly.
public function numericalGradient(string $name, float $eps = 1e-5): float
{
$keep = $this->w[$name];
$this->w[$name] = $keep + $eps; $up = $this->loss();
$this->w[$name] = $keep - $eps; $down = $this->loss();
$this->w[$name] = $keep;
return ($up - $down) / (2 * $eps);
}
public function step(float $lr): void
{
foreach ($this->gradients() as $name => $g) $this->w[$name] -= $lr * $g['value'];
}
}