COSMO · STUDY + TEACHING DECK

Computer Science & AI

July 29, 2026

A question worth sharing

Neural Networks as Functions

Build nonlinear models by composing affine transformations with activation functions.

ThinkCompareExplain

Today’s route

A question becomes an explanation.

Try

Enter through a problem. A partial attempt counts.

Connect

Use neuron, activation, layer to name what you notice.

Convince

Build an argument a partner can check.

The goal is a method someone else can use.

The opener

Start with a question.

Opening problem

For z=2x1-x2+0.5 with x=(1,3), what does a ReLU neuron output?

Think · 2 min

Sketch, test a small case, or make a first guess.

Pair · 3 min

Compare methods. Find one thing you agree on and one question.

Be ready to share your route, even if it is unfinished.

Compare before checking

Put two attempts in conversation.

Explain

Tell a partner how you started. Let them finish before responding.

Question

Which step feels certain? Which step still needs evidence?

Build

Combine one useful idea from each attempt into a new route.

Can you explain your partner’s method in your own words?

Board conversation

Notice before naming.

Look for structure

Try a small case and record every step.
Separate what changes from what stays fixed.
State the pattern in a sentence before naming the theorem.

Make a conjecture

We think ___ because ___. A case that might break this is ___.

What changed? What stayed the same?

Useful language

Now the language has a job.

neuron

weighted sum followed by an activation

activation

nonlinear function applied to a pre-activation

layer

collection of neurons evaluated together

Use one of these words to explain something you already noticed.

The core idea

Keep the idea. Explain the conditions.

\operatorname{ReLU}(z)=\max(0,z),\qquad z=w^\top x+b

The idea

Build nonlinear models by composing affine transformations with activation functions.

What to remember

A neural network is a chain of ordinary functions whose composition creates flexible behavior.

What information does this method keep?

Worked example

One route, one decision at a time

The problem

Evaluate h=ReLU(2x-1) and y=3h+2 at x=2.

Predict each next step

First compute 2(2)-1=3.
ReLU keeps the positive value, so h=3.
Then y=3(3)+2.
The output is 11.

Which step needs a reason, not just a calculation?

See the structure

The structure behind neural networks as functions

\operatorname{ReLU}(z)=\max(0,z),\qquad z=w^\top x+b

A neuron applies an affine rule followed by a nonlinearity; stacked affine rules alone collapse to one.

Read the representation

Build nonlinear models by composing affine transformations with activation functions.

Look together

Point to the part that explains the relationship. Change one input and predict what changes.

What can you see here that was harder to see in words?

Two approaches

Compare the reasoning.

Structural method

Name the objects and conditions.
Use a reusable relationship.
Explain why the method must work.

Brute force

May help with a small test case.
Can hide the general reason.
Becomes unreliable when the scale grows.

Which route can you justify? What would you test first?

Reasoning clinic

Find the step that needs repair.

Investigate these concerns

May help with a small test case.
Can hide the general reason.
Becomes unreliable when the scale grows.

Repair together

Name the missing condition. Try a small counterexample. Rewrite one step so it is justified.

A useful correction explains why the original step fails.

Work together

Investigate neural networks as functions

Your shared task

Groups trace values through a small network, then alter one weight and predict which intermediate computations change.

Explore

4 min - test cases

Compare

5 min - form a rule

Defend

4 min - defend the rule

Everyone should be able to explain the group’s final claim.

Share → connect

What did the room discover?

Bring your evidence

Share a conjecture, a useful example, or a route that changed.

Connections to investigate

A neural network is a chain of ordinary functions whose composition creates flexible behavior.
A complete solution names the condition that makes the method valid.
A counterexample is useful when it exposes a missing condition.

Which observation explains the mechanism?

Try → explain → check

Your turn to build the route

The problem

Show why two affine layers without activations are still one affine transformation.

Predict each next step

Write the layers as h=Ax+a and y=Bh+b.
Substitute h into the second equation.
Then y=(BA)x+(Ba+b).
The composition is affine, so no new nonlinear boundary appears.

Which step needs a reason, not just a calculation?

Convince a skeptic

Turn a pattern into a proof.

The claim

Nonlinear activations are what prevent stacked affine layers from collapsing into one affine map.

Build the argument

Affine maps are closed under composition.
An activation changes the functional form between layers.
The network can then create piecewise or curved decision behavior.

What does each step rely on? Where are the conditions used?

Stretch together

Take the idea somewhere new.

Transfer challenge

Construct a ReLU expression that is positive for 0<x<2 and zero outside this interval.

Choose a route

Name the objects, the conditions, and a relationship that might help. Start with one test case.

What transfers from the earlier example? What is different?

Peer review

Check the argument together.

Restate

Explain the other person’s route before evaluating it.

Test

Try a small case or boundary case. Does every step still make sense?

Revise

Fix an assumption, add a reason, or find a second method.

Where did your thinking change?

Beyond this problem

Where might this idea help?

A connection

Neural networks model images, language, signals, scientific systems, and high-dimensional decision rules.

Discuss the model

What objects or quantities are represented?
Which relationship from today is preserved?
What assumption would you check before trusting it?

Invent one more situation with the same structure.

Exit ticket

Leave with an explanation.

Question 1

For z=2x1-x2+0.5 with x=(1,3), what does a ReLU neuron output?

Question 2

Explain why this claim is true: Nonlinear activations are what prevent stacked affine layers from collapsing into one affine map.

Question 3

Construct a ReLU expression that is positive for 0<x<2 and zero outside this interval.

Choose one. Write an answer with a reason, then compare with a partner.

Reflect together

Make your thinking visible.

I used to think…

Name a claim or approach that changed.

Now I can explain…

Give the reason, not only the result.

I still wonder…

Write a question worth bringing to the next circle.

Share one sentence. Ask one follow-up question.

The idea to keep

A neural network is a chain of ordinary functions whose composition creates flexible behavior.

Use the matching practice sets to move from direct examples to proof and transfer.

A good question belongs to everyone in the room.