Want to create interactive content? It’s easy in Genially!

Get started free

Train your robot

San-Shan Huang

Created on November 28, 2025

Start designing with a free template

Discover more than 1500 professional designs like these:

Essential Quiz

Smart Quiz

Practical Quiz

Akihabara Quiz

Pixel Challenge

Math Calculations

Piñata Challenge

Transcript

Train your robot

Agent, you're training an AI model. Choose examples that teach it what a cat truly looks like. Pick well! Wrong data makes the platform unstable.

START

QUESTION 01 OF 05

Start Simple. Choose the example that actually teaches the model what a cat looks like.

QUESTION 01 OF 05

RIGHT!

Good start. You picked a clear example of a cat. The model begins learning.

NEXT QUESTION

OH NO!

This has no cat-like patterns. Noise weakens the model.

TRY AGAIN

QUESTION 02 OF 05

think wider. select the example that teaches the model more than it already knows.

QUESTION 02 OF 05

RIGHT!

Great choice. Variety helps the model understand cats from different viewpoints.

NEXT QUESTION

OH NO!

This pattern isn't real. Models can learn the wrong texture if the example is misleading.

TRY AGAIN

QUESTION 03 OF 05

some choices look close... but only one really helps. Spot the true cat pattern.

QUESTION 03 OF 05

RIGHT!

Nice. Mixing in cats of different colors and shapes helps the model understand that "cat" comes in many forms, not just one.

NEXT QUESTION

OH NO!

The shapes might be similar, but it's still wrong. The model starts mixing categories.

TRY AGAIN

QUESTION 04 OF 05

Every image shows a cat... but one offers a much cleaner signal than the rest.

QUESTION 04 OF 05

RIGHT!

Good call, Agent. You selected the clearest cat image, giving the model a solid pattern to lock onto.

NEXT QUESTION

OH NO!

You chose a real cat, but the details are hard for the model to pick up. Choose a clearer pattern.

TRY AGAIN

QUESTION 05 OF 05

edge cases matter. select the one that teaches the model to handle the unexpected.

QUESTION 05 OF 05

RIGHT!

Excellent. Showing edge cases strengthens the model and reduces bias.

NEXT

OH NO!

Similar texture, wrong species. This sample adds no real signal, only noise the model can’t use.

TRY AGAIN

RESULTS

CONGRATULATIONS!

Great work. Now that you've learned how the platform reacts to your choices, it's time to explore what happens when the model meets Digital City's data.