Reassessing pattern separation in the dentate gyrus

The dentate gyrus (DG) is postulated to be a “pattern separator” (Marr, 1971; Rolls, 1989a,b, 1990; Treves and Rolls, 1994). Yet, the definition of pattern separation has become a haze, with researchers using the term interchangeably to describe computational processes, changes in cell ensemble activity, and even behavioral phenomena (Leutgeb et al., 2007; McHugh et al., 2007; Clelland et al., 2009; Bakker et al., 2010). To accurately assess the large influx of papers purporting to attribute pattern separation to the DG, the concept must be reassessed; the original definition of pattern separation as a computational process and its newer, colloquial definition as a form of behavioral context discrimination must be accurately parsed if relations between the data are to be made.