arXiv - CSCL: "Deep Transformers without Shortcuts: Modifying Se…" - CleverLibre Social

arXiv - CSCL @arxiv_cscl@social.cleverlibre.org

Deep Transformers without Shortcuts: Modifying Self-attention for Faithful Signal Propagation. (arXiv:2302.10322v1 [cs.LG])

http://arxiv.org/abs/2302.10322 #arXiv #NLProc

Feb 22, 2023, 03:09 · · arxiv-cscl · · ·

Sign in to participate in the conversation