# Self-Attention **Domain:** Machine Learning / Transformer Architecture **Doc Type:** Concept **Maturity:** Developed **Related:** [[Attention Mechanism]], [[Transformer Models]], [[Representation Learning]], [[wiki/Bauhaus Architects of AI|Bauhaus Architects of AI]] ## Definition Self-attention computes relationships among elements within the same sequence, allowing each position to weight information from other positions. ## Bauhaus Architects Context The 2017 transformer architecture made self-attention its primary sequence-processing mechanism. ## Backlinks - [[wiki/Bauhaus Architects of AI|Bauhaus Architects of AI]] - [[articles/Bauhaus Architects of AI|Bauhaus Architects of AI]]