The Topos of Transformer Networks

arXiv:2403.18415v1 Announce Type: new
Abstract: The transformer neural network has significantly out-shined all other neural network architectures as the engine behind large language models. We provide a theoretical analysis of the expressivity of the transformer architecture through the lens of topos theory. From this viewpoint, we show that many common neural network architectures, such as the convolutional, recurrent and graph convolutional networks, can be embedded in a pretopos of piecewise-linear functions, but that the transformer necessarily lives in its topos …

abstract analysis architecture architectures arxiv cs.lg language language models large language large language models math.ct network networks neural network show theory through transformer transformer architecture type