In recent years, the neural manifold framework has spurred numerous advances in our understanding of cortical function. This framework proposes that the building blocks of neural computation are population-wide patterns of neuronal covariation, rather than the independent modulation of single neurons. Our workshop will bring together a diverse panel to discuss new advances and unanswered questions related to the identification and computational role of neural manifolds. The workshop will be split into two sessions. In the first, we will focus on theoretical and experimental work exploring the role of manifolds in neural computation. In the second session we will begin to merge the theory with mathematical and methodological considerations. We will explore the implications of linear vs nonlinear manifold estimation and statistical vs dynamical characterization of manifold activity. Each speaker will give a short talk, followed by a lengthy moderated discussion. This workshop is designed to inspire future work towards understanding manifold neural computation through experiments, theory, and computational techniques.