# Attention Mechanism **Domain:** Machine Learning / Natural Language Processing **Doc Type:** Concept **Maturity:** Developed **Related:** [[Functional Sentence Perspective]], [[Self-Attention]], [[Transformer Models]], [[wiki/Bauhaus Architects of AI|Bauhaus Architects of AI]] ## Definition An attention mechanism learns or computes weights that determine which elements of an input are most relevant to a current processing step. ## Bauhaus Architects Context Its comparison with functional sentence perspective is a noncausal structural isomorphism in relevance-weighted sequence processing. ## Backlinks - [[wiki/Bauhaus Architects of AI|Bauhaus Architects of AI]] - [[articles/Bauhaus Architects of AI|Bauhaus Architects of AI]]