Limits of population coding

摘要: To understand whether the population response expresses information better than aggregate of individual responses, sum contributions is frequently used as a baseline against which to assess population’s coding capabilities. Using processing theory, we show that this illusory: independent case theoretically impossible apply consistently any population. Instead, use noncooperative population, in each neuron processes common input independently others. transfer ratio, ratio Kullback-Leibler distances evaluated at and output measure ability, cooperative populations can perform either or worse baseline. Furthermore, effective only when poorly codes considered out context