Use when positioning an ICML submission against close ML literature, concurrent ICML submissions, recent public papers, workshop papers, ICLR/AISTATS/NeurIPS neighbors, and prior work under double-blind constraints.