Neural Net Quiz.

Test your understanding of neurons, weights, and network architecture.

This quiz covers

How neurons compute outputs
The role of weights, biases, and activation functions
How layers work together in a network
Key vocabulary from lessons 1-2

Recap

Neural network concepts at a glance.

Before you dive into the quiz, here is a quick visual recap of everything from Lessons 1 and 2. Think of a neural network as a factory assembly line: raw materials (data) enter on one end, get processed at each station (layer), and a finished product (prediction) comes out the other end.

The Neuron — a tiny decision-maker

Picture a judge at a talent show. Each performer (input) gets a score multiplied by how much the judge trusts their own taste in that genre (weight). The judge adds a personal bias — maybe they always lean generous — and then decides: does this act pass to the next round? That final yes/no decision is the activation function. Every neuron in a network does exactly this: weighted sum + bias + activation = output.

Weights — the learned knowledge

Weights are the numbers the network adjusts during training. Think of them as volume knobs on a mixing board. Some inputs get turned up loud (high weight = important), some get muted (low weight = unimportant), and some get inverted (negative weight = this input pushes the output down). Training is the process of finding the perfect setting for every knob.

Bias — the baseline mood

Without bias, a neuron with all-zero inputs always outputs zero. Bias is like a thermostat's default setting — it shifts the point at which the neuron "fires." A positive bias means the neuron is eager to activate; a negative bias makes it harder to trigger. This gives the network flexibility to fit patterns that do not pass through the origin.

Activation functions — the gatekeepers

Activation functions introduce curves into what would otherwise be a straight-line calculation. ReLU (the modern standard) is like a floor at zero: negative signals get silenced, positive signals pass through unchanged. Sigmoid squashes everything into a 0-to-1 range — perfect for probabilities. Without these gates, stacking layers would be pointless — the whole network would collapse into a single linear equation.

Layers — simple parts, complex whole

The input layer receives raw data (pixels, numbers, text). Hidden layers transform that data through learned patterns — first layer finds edges, second finds shapes, third finds objects. The output layer makes the final prediction. More layers means the network can learn more complex representations, but also needs more data and compute to train.

Here is the full flow visualized — data enters left, flows right, and a prediction emerges:

  DATA FLOW THROUGH A NEURAL NETWORK

  Raw Data          Pattern Detection        Decision
  ─────────         ──────────────────        ──────────
  pixels     ──▶    edges → shapes    ──▶     "cat" (92%)
  numbers    ──▶    trends → clusters ──▶     "buy" (78%)
  words      ──▶    syntax → meaning  ──▶     "positive" (85%)

  INPUT LAYER        HIDDEN LAYERS            OUTPUT LAYER
  (receives data)    (finds patterns)         (makes prediction)

  Each arrow = a weight (learned during training)
  Each node  = weighted sum + bias + activation
  Training   = adjusting ALL weights to reduce errors

The key insight: each neuron is embarrassingly simple — just multiply, add, and decide. But millions of these simple decisions, connected in layers, produce intelligence. That is the miracle of neural networks. Now let's test how well you understand each piece.

Part 1

Core concepts.

Training Recap

How a network learns from its mistakes.

Training is the process of adjusting weights and biases so the network gets better at its task. Here is the full loop, step by step:

1. Forward Pass — make a prediction

Data flows from input through hidden layers to the output. Each neuron computes its weighted sum + bias + activation. The final output is the network's prediction — maybe "92% cat, 8% dog." On the first try, this prediction is essentially random because the weights have not been trained yet.

2. Loss Calculation — measure the error

Compare the prediction to the correct answer using a loss function. If the network said "92% cat" but the image was a dog, the loss is high. If it said "95% dog," the loss is low. The loss function turns the error into a single number that the network can minimize.

3. Backpropagation — trace the blame

Work backwards from the output to figure out which weights contributed most to the error. Each weight gets a "blame score" (technically called a gradient) that says how much it should change and in which direction. Weights that contributed a lot to the error get adjusted more.

4. Weight Update — nudge toward correctness

Adjust every weight by a tiny amount in the direction that reduces the error. The size of the adjustment is controlled by the learning rate — too large and the network overshoots, too small and training takes forever. Then repeat: forward pass, loss, backprop, update. Millions of times.

That is the entire training loop. Forward pass (predict) → loss (measure error) → backpropagation (trace blame) → update (adjust weights) → repeat. Every AI model you have ever used — ChatGPT, Claude, Midjourney — learned through this exact process, billions of times over.

Part 2

Match the vocabulary.

Common Pitfalls

Mistakes beginners make about neural networks.

Before the final challenge, let's clear up the most common misconceptions about how neural networks work:

Myth: "More layers always means better"

Adding layers increases the network's capacity to learn complex patterns, but it also requires more training data and compute. A network that is too deep for the available data will overfit — it memorizes the training examples instead of learning general patterns. A 3-layer network trained well on enough data often beats a 100-layer network trained poorly.

Myth: "Neural networks understand like humans do"

A neural network that classifies cat photos does not "see" a cat the way you do. It detects statistical patterns in pixel values — edges, textures, shapes — that correlate with the label "cat." It has no concept of what a cat is, what it feels like to pet one, or that cats are alive. Pattern matching is powerful, but it is not understanding.

Myth: "Training data doesn't matter — the architecture does everything"

Architecture determines what the network can learn. Data determines what it does learn. A perfectly designed network trained on biased data will produce biased outputs. A network trained on too little data will overfit. A network trained on noisy, mislabeled data will learn noise. Data quality is at least as important as architecture quality.

Myth: "AI neurons work like brain neurons"

Artificial neurons were inspired by biological neurons, but they are radically simplified. A biological neuron uses electrochemistry, has timing-dependent behavior, and connects to about 7,000 other neurons on average. An artificial neuron is pure math: multiply, add, threshold. The inspiration was useful, but modern AI has diverged far from neuroscience.

Now you know the truth and the myths. The final section tests your understanding of the real mechanics — how neurons compute, how layers connect, and what makes networks powerful. Let's see what you've got.

Game Time

Collect the correct concepts.

Vocabulary

Your neural network glossary.

Keep this reference handy as you continue the course. These are the foundational terms that every AI concept builds on:

  NEURAL NETWORK GLOSSARY

  Term                 Definition
  ──────────           ──────────────────────────────────────
  Neuron               Weighted sum + bias + activation = output
  Weight               How much influence an input has (learned)
  Bias                 Baseline nudge — lets neuron fire at zero input
  Activation fn        Adds non-linearity (ReLU, Sigmoid, Step)
  Input layer          Receives raw data (pixels, numbers, tokens)
  Hidden layer         Finds patterns — edges, shapes, meanings
  Output layer         Makes the final prediction/decision
  Forward pass         Data flowing input → hidden → output
  Loss function        Measures how wrong the prediction was
  Backpropagation      Traces error back to each weight
  Gradient             How much (and which direction) to adjust a weight
  Learning rate        Size of each weight adjustment step
  Epoch                One full pass through all training data
  Overfitting          Memorizing data instead of learning patterns
  Softmax              Converts raw scores to probabilities (sum = 1)
  Parameters           Total weights + biases (GPT-4 ≈ 1.8 trillion)

You now have the vocabulary to read any AI article and understand what they are talking about. These terms come up in every course, every tutorial, every research paper. You are no longer on the outside looking in — you speak the language.

Next: Anatomy of a Prompt →