Abstract
Motor cortex (M1) has lateralized outputs, yet neurons can be active during movements of either arm. What is the nature and role of activity across the two hemispheres? We recorded muscles and neurons bilaterally while monkeys cycled with each arm. Most neurons were active during movement of either arm. Responses were strongly arm-dependent, raising two possibilities. First, population-level signals might differ depending on the arm used. Second, the same population-level signals might be present, but distributed differently across neurons. The data supported this second hypothesis. Muscle activity was accurately predicted by activity in either the ipsilateral or contralateral hemisphere. More generally, we failed to find signals unique to the contralateral hemisphere. Yet if signals are shared across hemispheres, how do they avoid impacting the wrong arm? We found that activity related to each arm occupies a distinct subspace, enabling muscle-activity decoders to naturally ignore signals related to the other arm.