What is reward sparsity and why is it a challenge in Q-Learning?

Reward sparsity occurs when feedback signals are rare or delayed.

Sparse rewards make it difficult for Q-learning to propagate useful gradients backward through time, slowing convergence and exploration.