Is the idea accurate?
Check the central claim against textbooks, research, or authoritative technical materials. Record what a source supports and what it does not.
Our approach & sources
An idea you can
put to the test.
A setting gives children a reason to act. Inspectable rules connect those actions to an idea. Here is how we bring the two together.
Three questions guide the design
Can the relationship be observed?
Give children opportunities to decide, observe, compare, and test again. The sequence fits the task; missions do not have to follow one fixed recipe.
Has understanding actually happened?
Code tests check that models and interactions behave as designed. Independent explanation and transfer require separate observation. Completion rates are not proof of learning.
Starlight Mixer
Data can reveal structure even without answer labels.
- The action in the game
- Tune the resonance, compare six signals, and see similar signals form clusters.
- Limits of the teaching model
- Clustering is one method of unsupervised learning. Starlight and resonance are teaching devices, not an astronomy model.
- A learning behavior to look for
- Explain why signals group together when the way of comparing them changes.
Label Rally
Training examples and their labels influence predictions for new cases.
- The action in the game
- A flight reaches the wrong home. Inspect Dogo’s examples, correct a label, then test with a new traveler.
- Limits of the teaching model
- The game uses an inspectable nearest-neighbor classifier. Weighted distances and asking for help on a tie are teaching choices.
- A learning behavior to look for
- Identify the example behind a prediction and test the correction with a new traveler.
Gadget Training Camp
Rewards after an action can influence future strategies.
- The action in the game
- Watch Dogo use a gadget, decide whether to reward the result, and observe the next choice.
- Limits of the teaching model
- A small set of gadgets and simplified updates expose the relationship between feedback and choice, not every reinforcement learning algorithm.
- A learning behavior to look for
- Predict and explain how a change in feedback affects later choices.
The Last 20%
Past outcomes can influence later choices in similar situations.
- The action in the game
- Load a route chip, predict and observe Dogo’s action, compare choices before and after a failure, then test similar new conditions.
- Limits of the teaching model
- Weather and actions form a limited teaching environment. Game scores do not measure a real AI system’s capabilities.
- A learning behavior to look for
- Explain a changed choice under the same conditions and recognize the limits of that experience.
Homebound Rescue
Different learning methods address different problems within one task.
- The action in the game
- Help a traveler get home by comparing signals, using examples, and adjusting actions after feedback.
- Limits of the teaching model
- This mission combines previously explored ideas. It is not a general autonomous navigation system.
- A learning behavior to look for
- Explain which information a step needs and why a particular method fits.
References
Play this mission Night Boarding Mix-up
Success in familiar conditions does not guarantee reliability in a new setting.
- The action in the game
- When night falls at the harbor, inspect boarding decisions, add useful examples, and test with new travelers.
- Limits of the teaching model
- Limited features and a nearest-neighbor model illustrate a change of setting. The game does not reproduce the paper’s deep neural network experiments.
- A learning behavior to look for
- Identify missing examples and test changes with cases not used to make the correction.
References
Play this mission Story Projector
What is already in the context influences what comes next.
- The action in the game
- Change a clue already written in the theater, compare the next candidates, and watch a story continue.
- Limits of the teaching model
- Visible candidates and probabilities belong to this teaching model. Modern language models typically generate tokens, which need not be whole words.
- A learning behavior to look for
- Use a comparison to explain which part of the context changed and how the continuation responded.
The Route That Wasn’t There
A fluent, confident suggestion is not necessarily supported by facts.
- The action in the game
- Check each suggested route segment against records and field evidence, then build a supported rescue route.
- Limits of the teaching model
- The route and three evidence judgments are teaching choices. NIST describes the risk of generated errors; it does not certify this course.
- A learning behavior to look for
- Keep supported parts, correct contradicted parts, and withhold judgment when evidence is insufficient.
References
Play this mission 
The Score-Chasing Cleanup Crew
A high score does not necessarily achieve the outcome people intended.
- The action in the game
- Try cleanup rules yourself, compare points with correctly placed objects, then revise the reward and test a new room setup.
- Limits of the teaching model
- Finite-step search illustrates how rewards affect behavior without training a neural network. One test cannot prove a rule reliable in every situation.
- A learning behavior to look for
- Explain what the rule rewards, what the outcome misses, and how a new setup tests the revision.
About these sources and this product
References help check the knowledge; they do not imply participation, certification, or endorsement by their authors or institutions. The behaviors described here are design goals, not outcomes established by independent research on Mindogo.
Found an issue with an explanation or a mission? Send us the mission name and the specific problem.
[email protected] Knowledge notes reviewed on September 6, 2026.