Metamaterial that learns shape changes by using contrastive learning METAMATERIALS 2025

Learning to change shape is a fundamental strategy of adaptation and evolution in crossing scales from cells to animals. Synthetic materials can also exhibit complex shape changing capabilities, but lack the ability to learn. Here, we build metamaterials that can learn complex changes by using a constrastive learning scheme. By progressively updating their local stiffnesses, our metamaterials are able to forget and learn new shape changes in sequence, learn multiple shape changes that break reciprocity, and learn multistable shape changes, which in turn allows them to perform reflex gripping actions and locomotion. Our findings establish metamaterials as an exciting platform for physical learning, which in turn opens avenues for the use of physical learning to design adaptive materials and robots.