Fanout Complexity of Symmetric Boolean Functions in
arXiv:2609.05153
Abstract
Whether can compute remains open. Computing is equivalent to implementing under reductions. This raises a more general question: for an arbitrary symmetric Boolean function , what fanout size is necessary and sufficient for computing in ? We show that the answer is exactly the transition radius : computing and implementing are equivalent under reductions. In particular, if for some constant , then computing is -complete. Combined with Paturi's theorem, our characterization implies that if , then any Boolean function in of approximate degree must be nonsymmetric.
arXiv commentcomments are welcome