Twisted M-theory and Holography

A few years ago, K.Costello proposed how to isolate a self-contained, protected subsector of the M2 brane AdS4/CFT3 correspondence and demonstrated the holographic duality explicitly for the OPE of the corresponding boundary local operators. The construction can be naturally extended to protected correlation functions, but several new challenges arise. I will discuss how to take the large N limit on the CFT3 side of the story.