Interactive Exploration of Hill Climbing Challenges
Click anywhere to place the agent and watch hill climbing in action!Click any scenario to see hill climbing behavior in different landscape features:
Start near the global peak and watch success!
Get stuck on a smaller peak - classic hill climbing problem!
Wander aimlessly on flat terrain with no direction!
Navigate a flat shelf that leads to higher ground!
Follow a narrow path where small wrong steps lead downward!
Hill climbing stops at the first peak it finds, even if taller peaks exist elsewhere. This is why the algorithm is called "greedy" - it takes the best immediate option without considering the bigger picture.
On flat terrain, hill climbing has no gradient to follow. All neighbors have the same utility, so the algorithm can't decide which direction leads to improvement.
Shoulders are flat regions that eventually lead upward. A patient algorithm that explores multiple moves can find the hidden upward path.
Ridges require precise navigation. One wrong step leads downward, making progress slow and requiring careful exploration of each move.