Fig. 8: Efficient resource allocation solutions considering RT costs.
a, Optimal allocation of neural resources in context Kacc considering different RT costs η. For reference, we have added the optimal solution of context Kacc in a RT-cost-free environment (blue dashed line). As the RT cost decreases, the allocation of resources appears to slightly flatten relative to the RT-cost-free solution. b, Decision bound as a function of η. As expected, the larger the RT cost, the smaller the decision bound. c,d, Same as a,b, but this time in context Krew. For reference, we have added the optimal solution of context Krew in an RT-cost-free environment (red dashed line). While the RT-cost solutions are similar to the RT-cost-free solution, contrary to the Kacc environment, the RT-cost solutions appear to get steeper as the RT cost decreases.
