A new study: How memory works – generative AI can help

Have you ever wondered how a complicated mechanism of human memory works?

Scientists have – so they've used a special generative AI tool to imitate human memory and research the patterns.

Let's find out more about what they've discovered.

A new discovery

Generative AI, a type of artificial intelligence, is helping researchers at UCL explore how memories work.

In this groundbreaking study, a cutting-edge generative neural network is employed to delve into the intricate processes of how brain networks learn and retain memories.

This innovative model emulates the collaborative workings of the hippocampus and neocortex, two key brain regions involved in memory, imagination, and planning.

Lead author Eleanor Spens highlights the recent strides in generative networks, offering insights into how they extract information from experiences, enabling us not only to recall the past but also to envision entirely new experiences.

The study posits that the act of replaying memories during moments of rest serves a crucial role in helping our brains discern patterns from past experiences, contributing to our ability to make predictions vital for survival.

How it works

During the research, the model underwent exposure to a vast array of 10,000 images portraying simple scenes.

The hippocampal network swiftly encoded each scene, while the neocortical network learned the intricate process of recreating scenes through patterns of activity.

This study provides valuable insights into the mechanisms behind the reconstruction of memories and helps elucidate the presence of "gist-like" distortions within them.

Previously, we talked about kids adapting to their kindergartens.

• Related news:
• research
• memory forming
• human brain
• technologies
• science
• 27 January 2024

---