Top Highlights

• We next trained neural-network models with different architectures and learning rules to perform the task. Networks that use the statistical properties of stimuli to enhance separability of the data via unsupervised learning during passive exposure provided the best account of the behavioral observations.
• These results provide key insights for the design of efficient training schedules that combine active learning and passive exposure in both natural and artificial systems.
• A large body of research has demonstrated that exposure to sounds early in life influences the ability to discriminate acoustic stimuli (Kuhl et al., 2003; Maye et al., 2002; Kral, 2013).
• In humans, previous work has demonstrated that, under specific conditions, interleaved passive exposure is beneficial for learning, sometimes to the extent that active sessions can be replaced with passive exposure and still yield similar performance (Wright et al., 2015).
• In other animals, which provide greater experimental access for investigating the neural mechanisms of learning, studies have focused mostly on the effects of perceptual learning (the experience-dependent enhancement in sensory discrimination) from exposure to stimuli during active training (Bao et al., 2004; Polley et al., 2006; Caras and Sanes, 2...

Domain