Physical sensorimotor contingencies facilitate learning in monkeys

Physical maze-solving with realistic objects enables learning that is orders of magnitude faster than in a comparable computerized virtual environment in rhesus monkeys.

• Why it matters: Virtual tasks can diverge from natural behavior, leaving uncertainty about whether they reveal the same learning computations and mechanisms as naturalistic settings.
• What they did: Rhesus monkeys were trained to solve a maze in two environments—one using physical objects and one using a computerized virtual setup—and learning timescales and generalization were compared.
• The result: Performance transfer between the physical and virtual settings was limited, implying that task environment strongly shapes learning and generalization in this model.

AI summary of the abstract · may contain errors · not medical advice · read the paper.