EulerFormer: Sequential User Behavior Modeling with Complex Vector Attention

arXiv:2403.17729v1 Announce Type: cross
Abstract: To capture user preference, transformer models have been widely applied to model sequential user behavior data. The core of transformer architecture lies in the self-attention mechanism, which computes the pairwise attention scores in a sequence. Due to the permutation-equivariant nature, positional encoding is used to enhance the attention between token representations. In this setting, the pairwise attention scores can be derived by both semantic difference and positional difference. However, prior studies often model the two …

abstract architecture arxiv attention behavior core cs.ir cs.lg data encoding lies modeling nature positional encoding self-attention transformer transformer architecture transformer models type vector