Ruth K. Broad Foundation Seminar on Neurobiology and Disease
Neurobiology welcomes Markus Meister, PhD, the Anne P. and Benjamin F. Biaggini Professor of Biological Sciences at the California Institute of Technology. Dr. Meister will present "Rapid learning of complex tasks: from phenomena to algorithms" in person, with a Zoom simulcast. Email email@example.com for details. ABSTRACT: Animals learn certain complex tasks remarkably fast, sometimes after a single experience. To bring such phenomena into the laboratory, we have studied the unconstrained behavior of mice in a labyrinth. The animal eagerly explores the new environment, making about 2000 turning decisions per hour. It quickly discovers the location of a reward in the maze and then navigates back to that location at will. It learns to make correct 10-bit choices after only 10 reward experiences. It also finds an error-free route back to the entrance on the first attempt. How can one explain such one-shot learning of complex actions in a novel environment? I will propose an algorithm that learns the structure of an arbitrary environment, discovers useful targets during exploration, and navigates back to those targets by the shortest path. It makes use of a behavioral module common to all motile animals, namely the ability to follow an odor gradient by trial and error. We show how the brain can generate internal "virtual odors" that serve as tags for locations of interest. This "endotaxis" algorithm can be implemented with a 3-layer neural circuit using only biologically realistic structures and learning rules. Components of this scheme are found in brains from insects to humans, suggesting that nature may have evolved a general mechanism for search and navigation on the ancient backbone of chemotaxis.