Revisiting mixture models of memory

Probabilistic mixture models have contributed significantly to advancements in visual working memory research in recent decades. In a new paper, Schurgin and colleagues revisit the basic assumptions of mixture models and suggest that we cannot understand memory without first considering perception.