NSF-BSF: early visual experience and the interplay between plasticity and stability in sensory learning

The National Science Foundation's Office of Multidisciplinary Activities awarded Trustees of Tufts College $766,569 on September 1, 2026, under the Social, Behavioral, and Economic Sciences program (CFDA 47.075) to investigate how biological and artificial systems balance plasticity and stability in sensory learning.

The research contrasts impaired sensory learning in humans whose vision was restored after early-life blindness with intact sensory learning in typically developed adolescents and adults. The project combines behavioral experimentation and computational modeling to test how altered visual experience in early life affects the interplay of plasticity and stability in visual and haptic sensory systems. Researchers implement different types of sensory learning—ranging from short-term adaptation within minutes to natural learning over years—in artificial neural networks and systematically manipulate those networks to identify principles underlying continuous learning without forgetting.

The study advances basic understanding of human learning while informing the design of more human-like artificial intelligence systems. The findings hold potential to improve prediction of how humans adapt to changing sensory demands, including those arising from aging or injury, and support the development of interactive technologies that require sensory learning for optimal use. The project provides interdisciplinary training in cognitive, developmental, and computational research for students and early-career researchers.

Performance occurs in Medford, Massachusetts, with a period of performance running through August 31, 2029.

Abstract

Human cognition and artificial intelligence face the same fundamental challenge: balancing plasticity (the ability to adapt to novel situations) and stability (the preservation of existing capabilities). Solving this challenge is essential for intelligent systems to learn continuously without forgetting. This project investigates how biological and artificial systems balance plasticity with long-term stability in the domain of sensory learning, which is a core foundation of cognition. The resulting knowledge advances basic understanding of human learning, while informing the design of more human-like AI systems. The identified principles hold potential to improve the ability to predict how humans cope with changing sensory demands, including those arising from aging or injury. The outcomes of this project are highly relevant to developers of interactive technologies as these technologies typically require sensory learning for optimal use. The project provides interdisciplinary training in cognitive, developmental, and computational research for students and early-career researchers, thereby supporting workforce development. To investigate how plasticity and stability interact to enable successful sensory learning, the researchers contrast impaired sensory learning in humans whose vision was restored after being blind in early life with intact sensory learning in typically developed adolescents and adults. This project combines behavioral experimentation and computational modeling techniques to test how altered visual experience in early life affects the interplay of plasticity and stability in visual and haptic sensory systems. Additionally, different types of sensory learning, ranging from short-term adaptation within minutes to natural learning over years, are implemented in artificial neural networks. Systematic manipulation of the artificial neural networks enables the researchers to identify the computational mechanisms underlying successful and impaired sensory learning in humans and machines. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the foundation's intellectual merit and broader impacts review criteria.- subawards are not planned for this award.