Uncertainty sampling has traditionally been the de facto, simplest acquisition function for active learning in semi-supervised learning. Comparatively cheap to compute and straightforward to interpret, uncertainty sampling has been known to suffer from myopic sampling bias that fails to properly explore the extent of geometric structure of the dataset prior to exploiting learning decision boundaries. As such, most work in active learning for graph-based learning has focused on the design of more intricate acquisition functions that are explorative in nature, though are almost always more costly to compute. We show that exploration and exploitation in graph-based semi-supervised learning can be achieved by ensuring that the underlying graph-based model reflects uncertainty at unlabeled nodes that in turn captures the underlying clustering structure of the dataset. We present theoretical results and empirical evidence to demonstrate that Poisson reweighted Laplacian learning, which has a well-defined continuum limit model, allows us to ensure exploration by simply applying a specific form of uncertainty sampling.