Blockwise Parallel Decoding for Deep Autoregressive Models
Mitchell Stern, Noam Shazeer, Jakob Uszkoreit
–Neural Information Processing Systems
To overcome this limitation, we propose a novel blockwise parallel decoding scheme in which we makepredictions for multiple time steps inparallel then back offtothe longest prefix validated byascoring model.
Neural Information Processing Systems
Feb-14-2026, 12:08:53 GMT