“…Beginning with Marr-Albus theory of the cerebellum in the 1970’s, expansion coding has been hypothesized to perform pattern separation. The ratio between sensory channels and expansion layer neurons, and the number of sensory inputs that individual expansion layer neurons receive, are theorized to be key parameters in balancing perception, discrimination, and generalization (Albus, 1971; Babadi and Sompolinsky, 2014; Cayco-Gajic and Silver, 2019; Hiratani and Latham, 2022; Ito, 1972; Jefferis et al, 2007; Jortner et al, 2007; Litwin-Kumar et al, 2017; Luo et al, 2010; Marr, 1969; Modi et al, 2020; Rajagopalan and Assisi, 2020). However, the perceptual and behavioral effects of altering hard-wired quantitative relationships have not been experimentally tested.…”