Parallelizing linear recurrent neural nets over sequence length

Publication
International Conference on Learning Representations (ICLR)