Formal models of memory based on temporally-varying representations

TL;DR

This paper reviews the development of formal models of memory that utilize temporally-varying internal states, highlighting their neural basis and applications in cognitive modeling across history and modern theories.

Contribution

It provides a comprehensive overview of the evolution of memory models based on scale-invariant temporal representations, connecting historical theories with current neural and cognitive models.

Findings

Neural phenomena support temporally-varying memory representations.

Modern models incorporate scale-invariant temporal history.

Connections established between formal models and memory task performance.

Abstract

The idea that memory behavior relies on a gradually-changing internal state has a long history in mathematical psychology. This chapter traces this line of thought from statistical learning theory in the 1950s, through distributed memory models in the latter part of the 20th century and early part of the 21st century through to modern models based on a scale-invariant temporal history. We discuss the neural phenomena consistent with this form of representation and sketch the kinds of cognitive models that can be constructed using it and connections with formal models of various memory tasks.