Back
ES EN
Interactive lab The AI Codex Interactive System
Initializing physics engine...
El Códice de la IA | RL 3D Playground
← Back to Index
psychology RL Playground
Train an AI agent using Q-Learning or a Deep Q-Network. Watch how it learns to reach the goal.
flag Goal: Green zone (+100)
warning Danger: Red zone (-50)
block Obstacle: Yellow zone (-30)
Espacio Play/Pause
R Reset
1-4 Change Environment
Control Panel
Agent
psychology
Q-Table Agent
Idle
Stats
0
Episode
0
Steps
0.0
Reward
0.0
Best
0
Q-States
0%
Success
Performance
Rewards Average (10)
Algorithm
Hyperparameters
Learning Rate 0.001
Discount (γ) 0.99
Exploration (ε) 0.20
Environment
route
Maze
motion_mode
Dynamic
view_in_ar
Obstacles
target
Simple
Visualization
Reward Design
1x
© 2025 The AI Codex Francisco Vidal Ureta / Principal Data Scientist & AI Strategist