Self-Attention Based Semantic Decomposition in Vector Symbolic Architectures

arXiv:2403.13218v1 Announce Type: new
Abstract: Vector Symbolic Architectures (VSAs) have emerged as a novel framework for enabling interpretable machine learning algorithms equipped with the ability to reason and explain their decision processes. The basic idea is to represent discrete information through high dimensional random vectors. Complex data structures can be built up with operations over vectors such as the "binding" operation involving element-wise vector multiplication, which associates data together. The reverse task of decomposing the associated elements is a combinatorially …

abstract algorithms architectures arxiv attention basic cs.ai cs.cv cs.sc data decision enabling framework information machine machine learning machine learning algorithms novel processes random reason self-attention semantic through type vector vectors