Clément Aubert, Proof circuits and parallel cut-elimination in logarithmic space



Terui, Mogbil and Rahli detailed how it was possible to compute Boolean functions in parallel with proof-nets for MLLu. We introduced Proof circuits to lighten the encoding of Boolean Circuits into proof-nets, extending previous results to sub-logarithmic classes of complexity. It happened that this way of constraining our proof-nets eased their simulation by Alternating Turing Machine, the most general kind of Turing Machine. We present a small extension to previous results, where proof-nets can be normalized with logarithmic space.