Cognitive Maps in Rats, Robots and Men: A Brain Inspired, Neuroevolutionary Approach

Abstract

In the mid twentieth century, Edward Tolman wrote an influential paper, “Cognitive Maps in Rats and Men”, to describe the flexible, intelligent behavior observed in animals. In a series of seminal behavior experiments, Tolman demonstrated that rats have cognitive mapping characteristics, that is, they can deliberate, adapt to change, form hypotheses, and show imagination. It should be noted that these capabilities performed by rats most definitely outperform current man made systems. To better understand cognitive maps in rats and humans, and to build an artificial system with these capabilities, we propose a unique convergence of experimental and computational modeling techniques. Our approach will: (i) observe and manipulate circuits from multiple brain areas simultaneously, all during cognitive mapping experiments in the rodent; (ii) optimize large scale parameter spaces through machine learning and coevolutionary computation to match these neurophysiological responses and behavior in spiking neural networks; (iii) port the spiking neural networks onto embedded, non Von Neumann neuromorphic chips to demonstrate cognitive mapping on power efficient autonomous robots. A brain inspired system that could demonstrate all of these traits would be a landmark achievement and would lead to a deeper understanding of how the brain represents and expresses cognitive behavior.

Document Details

Document Type
DoD Grant Award
Publication Date
Jan 14, 2022
Source ID
FA95501910306

Entities

People

  • Jeffrey L. Krichmar

Organizations

  • Air Force Office of Scientific Research
  • United States Air Force
  • University of California, Irvine

Tags

Readers

  • Military History of the United States in the 20th Century.
  • Neural Network Machine Learning.
  • Team-Based Human-Centered Cognitive Task Decision Making and Information Performance.

Technology Areas

  • AI & ML
  • AI & ML - Autonomous Systems
  • Autonomy
  • Space