In GR we have a metric space given by the metric $g^{\mu\nu}(x)$. This allows us to deetermine the curvature $R^{\mu\nu}$ at each point in space-time.
This curved 4D manifold is not required to be embedded in some higher dimensional space. But it is perfectly possible to do so. The Nash embedding theorems and related theorems say that an Einsteinian manifold can be embedded in a flat space given enough dimensions (probably provided it is simply connected). Let $\phi^N$ be the coordinates of some higher dimensional space. In which case we could write the metric as:
$$g_{\mu\nu}(x) = \sum_N \partial_\mu \phi^N(x) \partial_\nu \phi^N(x)$$
Where $N$ ranges over a suitably high enough number. If we substitute this back into the GR equations, nothing would change except we could solve for $\phi^N$ (where there would be some freedom of choice).
But where it would become interesting is if the fields $\phi^N(x)$ were observable by themselves. e.g. had an action (although it looks like the action would cancel itself out in a way):
$$\sqrt{g} (g^{\mu \nu}(x) \partial_\mu \phi^N(x) \partial_\nu \phi^N(x) + m^2 \phi^N(x)\phi^N(x)) = \sqrt{g} (1 + m^2) \phi^N(x)\phi^N(x) $$
Although perhaps there would be better actions for the $\phi$ fields maybe if they were non-abelian.
But anyway, I have heard about ideas that the Universe we live in is a 4-dimensional membrane in a higher dimensional space but I don't know if the above math is how that is realised. The question is, are there theories of gravity such that the metric desrcibes the embedding of a manifold in a higher dimensional space? And what would be the evidence for or against such an idea?