Brain Challenge Zone

Train your brain with challenges inspired by AI and neural networks. Each puzzle helps develop critical thinking skills essential in the age of artificial intelligence.

Agilist Brain Challenge Zone

Cognitive Learning

Number Tower Challenge
Game 1
Easy
FilledEmptyEmpty

Number Tower Challenge

Challenge your brain with this mathematical puzzle that mimics how neural networks process information layer by layer.

📋 Instructions

Instruction 1
Fill in all missing numbers in the tower by applying the rule of how signals combine — just like neurons in the brain or layers in an AI network.
Instruction 2
Start from the bottom, top or middle row — these are your input values. Each number above equals the sum of the two numbers directly below it.
Instruction 3
Train numerical reasoning and pattern recognition — key functions of both the human brain and artificial intelligence. Understand how neurons combine signals layer by layer (like inputs in a neural network).
Download Puzzle & Solution
Solution preview
Synapse Matrix
Game 2
Easy
FilledEmptyEmpty

Synapse Matrix

Decode the pattern recognition logic of neural networks. This puzzle simulates how error-detection neurons fire when they encounter unexpected patterns, teaching you how AI systems identify anomalies and imbalances in data.

📋 Instructions

Instruction 1
Fill in the missing icons in the fourth row using the rule. This exercise is similar to error detection neurons, which activate when an unexpected combination occurs.
Instruction 2
  • Observe three rows of inputs (A, B, C).
  • For each column, count how many appear.
  • If the number is odd → ⚡, even → 💤.
Instruction 3
This is a 3-input neuron:
  • It fires when the pattern is unbalanced.
  • It stays quiet when the system is in harmony.
Download Puzzle & Solution
Solution preview
Neural Fusion
Game 3
Moderate
FilledFilledEmpty

Neural Fusion

Experience how multiple neurons work together to make a single decision. This challenge mirrors the way real neural networks combine weighted inputs from different sources to produce a unified output, fundamental to deep learning.

📋 Instructions

Instruction 1
Calculate the activation of three neurons using weighted inputs, then combine their results to find the Fusion Score and see if the whole network fires.
Instruction 2
  • Calculate each neuron's activation and compare with thresholds. If the activation score is above the threshold, the neuron fires (⚡); otherwise, it remains inactive (💤).
  • Compute the Fusion Score and compare with fusion threshold. If the Fusion Score is above the Fusion Threshold, the entire network fires (⚡); otherwise, it stays inactive (💤).
Instruction 3
This exercise mirrors how real neurons integrate multiple signals—each with different strengths—and how the brain fuses them into a single decision or perception.
Download Puzzle & Solution
Solution preview
Gradient Descent
Game 4
Expert
FilledFilledFilled

Gradient Descent

Master the core learning mechanism of AI. This advanced puzzle demonstrates how neural networks learn by iteratively adjusting weights to minimize error, the same process that powers everything from ChatGPT to self-driving cars.

📋 Instructions

Instruction 1
Reduce the total error between predicted and target outputs by iteratively adjusting the weights until the model learns.
Instruction 2
  • Start with given weights.
  • Calculate predictions (P) and errors (t).
  • Calculate MSE using the provided formula.
  • Adjust weights logically to reduce the error.
  • Repeat until error is minimal or network stabilizes.
Instruction 3
This exercise mirrors how learning happens in both AI and the brain. Connections (weights) strengthen or weaken based on feedback, minimizing error through repeated adaptation.
Download Puzzle & Solution
Solution preview

AI Governance Lab

AI Compliance Quest
Game 5
Easy
FilledEmptyEmpty

AI Compliance Quest

Evaluate structured AI controls, calculate compliance points, and determine whether the system satisfies the required conditions.

📋 Instructions

Instruction 1
Status of the controls:
  • K1 = Full
  • K2 = Partial
  • K3 = Not met
  • K4 = Full
  • K5 = Partial
Rules are:
  • Full = 100%
  • Partial = 50%
  • Not met = 0%s
Conditions are:
  • K3 is mandatory.
  • Minimum required: 80 point.
Instruction 2
  • Calculate points per control and total points.
  • Check conditions!
Instruction 3
Understand how AI compliance is evaluated through structured controls and criteria.
Download Puzzle & Solution
Solution preview
AI Risk Lab
Game 6
Moderate
FilledFilledEmpty

AI Risk Lab

Assess initial AI risks, apply mitigating controls, and verify whether any critical residual risk keeps the system out of compliance.

📋 Instructions

Instruction 1
Existing controls:
  • R1 → reduces risk 1 by 40%
  • R2 → no control for risk 2
  • R3 → reduces risk 3 by 50%
Threshold is:
  • Risk ≥ 15 = Critical
Rule is:
  • System is NOT compliant if any critical risk remains after controls are applied and threshold is exceeded.
Instruction 2
  • Calculate initial risk scores.
  • Apply controls and calculate reduced scores..
  • Check threshold.
  • Check rules!
Instruction 3
Understand how AI risks are assessed and mitigated using structured evaluation methods.
Download Puzzle & Solution
Solution preview
AI Fairness Check
Game 7
Expert
FilledFilledFilled

AI Fairness Check

Compare approval rates across multiple groups and verify whether the model meets a defined fairness threshold.

📋 Instructions

Instruction 1
  • Fairness must be evaluated across all groups, not just one comparison.
Fairness Condition:
  • The difference in approval rates between ANY two groups must not exceed 10%
Instruction 2
  • Calculate approval rates.
  • Check pairwise differences.
  • Hit for differences (Group A - Group B).
  • Apply Fairness condition.
Instruction 3
Understand how to measure and compare fairness across different groups using data-driven metrics.
Download Puzzle & Solution
Solution preview