# Context-sensitive neocortical neurons transform the effectiveness and efficiency of neural information processing

**Authors:** Ahsan Adeel, Mario Franco, Mohsin Raza, Khubaib Ahmed  
**arXiv:** [2207.07338](https://arxiv.org/abs/2207.07338)  
**Date:** 2022-07-15  
**Summary page:** https://www.summarizepaper.com/en/arxiv-id/2207.07338v1/

## Key points

- The study explores the mechanism by which context-sensitive neocortical neurons use their apical inputs to amplify coherent feedforward (FF) inputs, and how this mechanism can provide useful neural computation.
- A network composed of such local processors restricts the transmission of conflicting information to higher levels, greatly reducing the amount of activity required to process large amounts of heterogeneous real-world data.
- When processing audiovisual speech, these local processors use seen lip movements to selectively amplify FF transmission of the auditory information that those movements generate and vice versa.
- This form of neural information processing is well-matched to the abilities of mammalian neocortex in terms of its processing and learning capabilities.
- The study shows that this mechanism is far more effective and efficient than the best available forms of deep neural nets.
- This offers a step-change in understanding the brain's mysterious energy-saving mechanism and inspiring advances in designing enhanced forms of biologically plausible machine learning algorithms.

## Summary

The study titled "Context-sensitive neocortical neurons transform the effectiveness and efficiency of neural information processing" explores the mechanism by which context-sensitive neocortical neurons use their apical inputs to amplify coherent feedforward (FF) inputs, and how this mechanism can provide useful neural computation. The researchers demonstrate that a network composed of such local processors restricts the transmission of conflicting information to higher levels, greatly reducing the amount of activity required to process large amounts of heterogeneous real-world data. For instance, when processing audiovisual speech, these local processors use seen lip movements to selectively amplify FF transmission of the auditory information that those movements generate and vice versa. This form of neural information processing is well-matched to the abilities of mammalian neocortex in terms of its processing and learning capabilities. The study shows that this mechanism is far more effective and efficient than the best available forms of deep neural nets, offering a step-change in understanding the brain's mysterious energy-saving mechanism and inspiring advances in designing enhanced forms of biologically plausible machine learning algorithms. The authors are Ahsan Adeel, Mario Franco, Mohsin Raza, and Khubaib Ahmed.