
via Baylor College of Medicine
UMass Amherst, Baylor researchers collaborate on imitating brain’s ‘replay’ ability
Artificial intelligence (AI) experts at the University of Massachusetts Amherst and the Baylor College of Medicine report that they have successfully addressed what they call a “major, long-standing obstacle to increasing AI capabilities” by drawing inspiration from a human brain memory mechanism known as “replay.”
First author and postdoctoral researcher Gido van de Ven and principal investigator Andreas Tolias at Baylor, with Hava Siegelmann at UMass Amherst, write in Nature Communications that they have developed a new method to protect – “surprisingly efficiently” – deep neural networks from “catastrophic forgetting” – upon learning new lessons, the networks forget what they had learned before.
Siegelmann and colleagues point out that deep neural networks are the main drivers behind recent AI advances, but progress is held back by this forgetting.
They write, “One solution would be to store previously encountered examples and revisit them when learning something new. Although such ‘replay’ or ‘rehearsal’ solves catastrophic forgetting,” they add, “constantly retraining on all previously learned tasks is highly inefficient and the amount of data that would have to be stored becomes unmanageable quickly.”
Unlike AI neural networks, humans are able to continuously accumulate information throughout their life, building on earlier lessons. An important mechanism in the brain believed to protect memories against forgetting is the replay of neuronal activity patterns representing those memories, they explain.
Siegelmann says the team’s major insight is in “recognizing that replay in the brain does not store data.” Rather, “the brain generates representations of memories at a high, more abstract level with no need to generate detailed memories.” Inspired by this, she and colleagues created an artificial brain-like replay, in which no data is stored. Instead, like the brain, the network generates high-level representations of what it has seen before.
The “abstract generative brain replay” proved extremely efficient, and the team showed that replaying just a few generated representations is sufficient to remember older memories while learning new ones. Generative replay not only prevents catastrophic forgetting and provides a new, more streamlined path for system learning, it allows the system to generalize learning from one situation to another, they state.
For example, “if our network with generative replay first learns to separate cats from dogs, and then to separate bears from foxes, it will also tell cats from foxes without specifically being trained to do so. And notably, the more the system learns, the better it becomes at learning new tasks,” says van de Ven.
He and colleagues write, “We propose a new, brain-inspired variant of replay in which internal or hidden representations are replayed that are generated by the network’s own, context-modulated feedback connections. Our method achieves state-of-the-art performance on challenging continual learning benchmarks without storing data, and it provides a novel model for abstract level replay in the brain.”
Van de Ven says, “Our method makes several interesting predictions about the way replay might contribute to memory consolidation in the brain. We are already running an experiment to test some of these predictions.”
The Latest Updates from Bing News & Google News
Go deeper with Bing News on:
AI capabilities
- All In On AI: How Smart Companies Win Big With Artificial Intelligence
AI has been hitting the headlines recently, with generative AI, in particular, generating a great deal of interest. Two tools - the large language model chatbot ChatGPT and image generator Dall-E - ...
- Accelerating AI for growth: The key role of infrastructure
High-performance AI infrastructure offers CIOs and enterprises a way to cost-effectively manage and accelerate growth of production AI.
- Jurny Selects RAD AI for Creative Intelligence and Influencer Marketing
LOS ANGELES, January 26, 2023--Rad Technologies, Inc., doing business as Rad Intelligence, an artificial intelligence company that automates ROI-based creative intelligence, announced today that they ...
- Pentagon’s AI chief says data labeling is key to win race with China
"If we're going to beat China, and we have to beat China in AI," CDAO Craig Martell said, "we have to find a way to label at scale." ...
- Microsoft's AI push has some analysts raving about the stock even after lackluster earnings
Microsoft 's earnings report earlier this week was largely considered dissatisfying to analysts. But the tech giant's focus on artificial intelligence, or AI, is keeping many on Wall Street optimistic ...
Go deeper with Google Headlines on:
AI capabilities
[google_news title=”” keyword=”AI capabilities” num_posts=”5″ blurb_length=”0″ show_thumb=”left”]
Go deeper with Bing News on:
Deep neural networks
- Deepfakes: Faces created by AI now look more real than genuine photos
Even if you think you are good at analyzing faces, research shows many people cannot reliably distinguish between photos of real faces and images that have been computer-generated. This is ...
- MS and depression: Could damage to the same neural circuit be a likely cause?
Scientists have found that multiple sclerosis lesions affect similar areas of the brain as depression, which suggests transcranial magnetic stimulation might be an effective treatment.
- AI can ‘see’ people through walls using WiFi signals
A team at Carnegie Mellon University developed a deep neural network to digitally map human bodies when in the presence of WiFi signals. The researchers said they developed the technology in an effort ...
- Painting with mysterious origins is likely a Raphael, according to AI study
The technology found the de Brécy Tondo, which features a woman and baby, closely matched Raphael's Sistine Madonna.
- Artificial intelligence study determined a painting with mysterious origins is likely a Raphael, researchers say
The technology found the de Brécy Tondo, which features a woman and baby, closely matched Raphael's Sistine Madonna.
Go deeper with Google Headlines on:
Deep neural networks
[google_news title=”” keyword=”deep neural networks” num_posts=”5″ blurb_length=”0″ show_thumb=”left”]