# Questions on 2016 Paper: Why Neurons Have Thousands of Synapses

**URL:** <https://discourse.numenta.org/t/questions-on-2016-paper-why-neurons-have-thousands-of-synapses/7912>\
**Category:** Numenta Theory\
**Created:** [October 1, 2020, 10:19am UTC](https://discourse.numenta.org/t/questions-on-2016-paper-why-neurons-have-thousands-of-synapses/7912 "2020-10-01T10:19:12Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![TaherHabib](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.numenta.org/taherhabib/32/2895_2.png) [@TaherHabib](https://discourse.numenta.org/u/TaherHabib)\
**Post date:** [October 1, 2020, 10:19am UTC](https://discourse.numenta.org/t/questions-on-2016-paper-why-neurons-have-thousands-of-synapses/7912/1 "2020-10-01T10:19:12Z")

</div>

Hello,

Based on the Implementation details given in the [titled](https://www.frontiersin.org/articles/10.3389/fncir.2016.00023/full#h6) paper (and also in [2016’s Continuous Online Sequence Learning Paper](https://arxiv.org/abs/1512.05463)), I have the following two questions:

1. In equation (4) of the titled paper, what happens if there are _more than one cells predicted in a **single** _ mini-column? Will all of the predicted cell’s active dendrites receive a reinforcement of their synapses connected to active pre-synaptic cells? If yes, then does this not negatively effect the storage efficiency of the cells of the network?  
Equation (3) does not put any columnar inhibition on prediction of a cell, only equation (2) puts a columnar inhibition on the activity of the cell.

2. On a similar note to the first one, what happens if – in equation (5) – there are _more than one dendritic segments_ (from all the dendritic segments on all the cells in a particular **unpredicted** mini-column) that have a maximum overlap with the previous time-step’s activity pattern?

---

<div class="post-metadata">

**Author:** ![AMZ](https://avatars.discourse-cdn.com/v4/letter/a/65b543/32.png) [@AMZ](https://discourse.numenta.org/u/AMZ)\
**Post date:** [October 2, 2020, 4:15am UTC](https://discourse.numenta.org/t/questions-on-2016-paper-why-neurons-have-thousands-of-synapses/7912/2 "2020-10-02T04:15:59Z")

</div>

Hi,

1- All the predictive cells within the active mini-column become active. However, the reinforcement of the synaptic connections can be applied only to the learning cells.

> If yes, then does this not negatively effect the storage efficiency of the cells of the network?

There is always a likelihood of activating two or more cells within the same mini-column, but it is very low! Thus, I would say ‘yes’ it negatively affects the storage efficiency, but it is negligible.

2- If you have more than one active dendritic segment within a cell, that cell becomes predictive. During learning, all the active dendritic segments will be updated (assuming the cell status is changed from predictive to active.).

---

<div class="post-metadata">

**Author:** ![TaherHabib](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.numenta.org/taherhabib/32/2895_2.png) [@TaherHabib](https://discourse.numenta.org/u/TaherHabib)\
**Post date:** [October 2, 2020, 9:14am UTC](https://discourse.numenta.org/t/questions-on-2016-paper-why-neurons-have-thousands-of-synapses/7912/3 "2020-10-02T09:14:47Z")

</div>

Thank you for your response!  
I’ll implement this in my ongoing HTM implementation for learning of Simple and Embedded Reber Grammars and see how the network evolves. 🙂