“…The accumulation process laid out here and shown in Figure 2 can be mimicked by a correlated multiple-accumulator LBA (see Kvam, 2019a, , Figure 8), which can be implemented by a simple neural circuit (Tajima et al, 2019, , supplementary note 3). Certainly, the model can be informed by a better understanding of the similarity relations between the available responses, which may suggest that (for example) edge categories are more distinct or that the underlying construct does not map onto a purely unidimensional representation (Dodds et al, 2012;Kvam & Turner, 2021). Converging evidence from behavioral approaches like multidimensional scaling (Shepard, 1962) and neural similarity measures like representational similarity analysis or neural decoding rates (Kriegeskorte et al, 2008;Raizada & Connolly, 2012) should shed light on better ways to translate physical stimuli into psychological representations that can be incorporated into a model like ours.…”