- Case
- P
- Dry D
- 1
- Hot H
- 0
Nodes weights and outputs
Year 7 · Lesson 2 · Content 2026.11.02.1
All supplied-record lessons · Open interactive lesson
Supplied-record alternative. Enter your responses in your school assignment. These supplied results support learning but do not show that you ran an algorithm.
Your goal
Explain why the same clues can give a different answer when one clue has more influence.
To be able to explain how changing weights can change a neural network output.
Before you start
If dry = 1, does 1 multiplied by 2 equal 2 or 3?
Choices: 2 / 3
Bridge feedback after trying
1 × 2 = 2. Think of one group of two. 0 multiplied by any number is 0.
Words for this lesson
- Node
- A unit that performs a calculation.
- Weight
- A number that changes a clue's influence.
- Output
- The result produced by the model.
The idea
A network contains connected calculation units called nodes. Each connection has a weight. A weight changes how much an input contributes. Our small example adds two weighted inputs and compares the total with a cut off. A real network can contain many nodes and different calculation rules. Training can adjust weights and added numbers called biases.
A completed example
Practice only: a light sensor gives 1 and its weight is 4. Its contribution is 1 x 4 = 4. If a second sensor gives 0 with weight 2, its contribution is 0. Total 4. With a cut off of 5, this practice model outputs Wait.
Check your understanding
A weight changes a clue's ...
- Influence
- True label
- Colour
Practice feedback after trying
Influence. It multiplies the input contribution.
Can a score change while the output stays Check?
- Yes
- No
- Only if the input changes
Practice feedback after trying
Yes. Two different totals can both meet the cut off.
What may training adjust?
- Weights and biases
- The true test labels
- Only the diagram colour
Practice feedback after trying
Weights and biases. These are numerical model settings.
Task 1 · Make a prediction
Prediction
Options: Wait / Check
Write your prediction, or state that you skipped it.
Try it using supplied records
- Select plant P. Predict Wait or Check with weight set A.
- Press Calculate. Reveal each product, then the total, then the output.
- Switch only the dry weight from 2 to 3. Predict again and compare.
- Use Q, R and S to check whether a changed weight always changes the output.
Manual weight experiment
Complete data and source records
| Case | Dry D | Hot H |
|---|---|---|
| P | 1 | 0 |
| Q | 0 | 1 |
| R | 1 | 1 |
| S | 0 | 0 |
Read as stacked records
- Case
- Q
- Dry D
- 0
- Hot H
- 1
- Case
- R
- Dry D
- 1
- Hot H
- 1
- Case
- S
- Dry D
- 0
- Hot H
- 0
| Set | Dry weight | Hot weight | Bias | Check at score |
|---|---|---|---|---|
| A | 2 | 1 | 0 | 3 or more |
| B | 3 | 1 | 0 | 3 or more |
Read as stacked records
- Set
- A
- Dry weight
- 2
- Hot weight
- 1
- Bias
- 0
- Check at score
- 3 or more
- Set
- B
- Dry weight
- 3
- Hot weight
- 1
- Bias
- 0
- Check at score
- 3 or more
Expected results · supplied
These are supplied reference results. They are separate from your own explanation. No code needs to run to read them.
1. score = D*dryWeight + H*hotWeight + 0. Output Check if score >= 3, otherwise Wait. Set A scores 2,1,3,0 and outputs Wait,Wait,Check,Wait. Set B scores 3,1,4,0 and outputs Check,Wait,Check,Wait.
2. Only dry weight is editable through buttons Set A and Set B. Inputs are chosen through named case buttons. A prediction is required before reveal. Changing the setting clears the current reveal and preserves the previous comparison.
3. Use an accessible left-to-right diagram with textual equivalents for each input, weight, node total and output. No loss, epoch, activation selector or hidden layer controls. Label the interaction Manual weight experiment.
Help with this step
Hint 1
Look at the dry input and its weight.
Hint 2
Multiply each input by its weight. Add the two contributions. A total of exactly 3 gives Check.
Hint 3
Practice: input 1 with weight 5 contributes 5; input 0 with weight 9 contributes 0. Changing a weight attached to zero still contributes zero.
Task 2 · Use the evidence
- Record the scores and outputs for P with both settings. Choose one other case and complete its comparison.
- Explain what changed and what stayed the same for P.
- Name a limit of this one-node model. All four cases remain available for practice and the summary shows the complete comparison.
My explanation
Write in your school assignment. Keep these answers in your own words.
P comparison
- A score
- A output: Wait / Check
- B score
- B output: Wait / Check
Other case
- Case: Q / R / S
- A score
- A output: Wait / Check
- B score
- B output: Wait / Check
Your explanation
- What changed and what stayed the same
- One model limit
Explain a fresh example
A dry input stays at 1 while its weight rises from 2 to 3. The cut off stays 3. Explain why the output changes.
My confidence
I need another explanation / I can explain with help / I can explain with my own example
My next step
Write what you will do next.
Optional depth
Why does changing the hot weight have no effect on P? Why does R keep the same output even though its score changes?
What this activity can show
The pupil changes a setting manually. This illustrates a mechanism, not evidence that software learned these weights. Bias is fixed at zero; hidden layers and activation functions are not modelled.
This is a student learning record. Submission is completed through the school assignment system.