# From Piecewise Linear Networks to Context Selected Linear Mappings

**URL:** <https://discourse.numenta.org/t/from-piecewise-linear-networks-to-context-selected-linear-mappings/12520>\
**Category:** Lounge\
**Created:** [August 1, 2026, 10:07pm UTC](https://discourse.numenta.org/t/from-piecewise-linear-networks-to-context-selected-linear-mappings/12520 "2026-08-01T22:07:45Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![SeanOConnor](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.numenta.org/seanoconnor/32/3441_2.png) [@SeanOConnor](https://discourse.numenta.org/u/SeanOConnor)\
**Post date:** [August 1, 2026, 10:07pm UTC](https://discourse.numenta.org/t/from-piecewise-linear-networks-to-context-selected-linear-mappings/12520/1 "2026-08-01T22:07:45Z")

</div>

It is widely understood that ReLU neural networks are piece-wise linear mappings.  
I have tried to give an ‘almost no choice’ argument leading to a layer-wise context mechanism:  
[From Piecewise Linear Networks to Context Selected Linear Mappings](https://from%20piecewise%20linear%20networks%20to%20contextselected%20linear%20mappings)