Attention Mechanism

A computation that lets a model weigh how much each input token should influence the output for a given position. Self-attention compares every token to every other token, then mixes their representations using those weights. This is the core operation inside a transformer, and it explains both the strength and the cost of the architecture.

Definición editorial de DecipherU. Para definiciones técnicas autorizadas, consulta NIST AI RMF, ISO/IEC 42001, o las publicaciones oficiales de los laboratorios de IA.