Panagiotis Parthenios Sakagiannis, Anna-Maria Jürgensen, Martin Paul Nawrot
A three-layered modular architecture unifies diverse computational models of Drosophila larval behavior, enabling realistic closed-loop simulations from locomotion through chemotaxis to associative learning.
Monkeys generalize many visual categorization rules, such as animate versus inanimate, but fail on culturally defined ones, placing their behavior closer to networks trained on images alone than to humans.
