Congreso CIMED I Congreso Internacional de Museos y Estrategias Digitales UPV, 25 y 26 de Marzo 2021 Doi: https://dx.doi.org/10.4995/ CIMED21.2021. 12487 2021, Editorial Universitat Politècnica de València Postcatalog, or how and why to physically experience a high- dimensional vector space CIMED AWARDS English Version, Versión en Español en índice Nuria Rodríguez-Ortega a , Gabriel Ramírez Gonzálvez b , Javier de la Rosa c , María Ortiz Tello d y Leticia Crespillo Marí e a Universidad de Málaga (nro@uma.es), b Opossum Studios (Gabriel@opossum.es), c UNED (versae@gmail.com), d Universidad de Málaga (mortiztello@uma.es) y e Universidad de Málaga (lcrespillom@uma.es) Abstract The immersive installation Postcatalog, developed ad hoc for the exhibition Unchained Catalogs organized by the Vice-Rectorate of Culture of the University of Malaga (12-17-2020, 01-29-2021), addresses two lines of inquiry. Firstly, the project delves into how to make high-dimensional spaces derived from computational image processing physically experiential. Secondly, the project raises the question of how traditional catalog ordering will take shape in a media ecology where the focus on books has recently shifted toward new scenarios of mixed realities that include viewers as active entities in their ambivalent cognitive-emotional dimension. The well-known diagram conceived by Alfred H. Barr for the cover of the Cubism and Abstract Art exhibition catalog (MOMA, 1936) was taken as the basis of our inquiry. In this diagram, Alfred H Barr represented the evolution of the artistic movements that had emerged during the first decades of the 20th century as genealogical relationships following a teleological conception. In the first phase of the project, the diagram was transformed into a high-dimensional space where the artworks corresponding to the isms and poetics represented in the diagram were distributed in clusters according to the visual similarities detected by an Inception convolutional neural network (CNN). They were then represented in the two-dimensional space generated by the UMAP distribution algorithm. In this way, the distance between images – transformed into numerical vectors – is translated into mathematically computed degrees of visual similarity. The result of this first experiment was the Barr X Inception CNN project. In the second phase, the two-dimensional space generated by the UMAP distribution algorithm was transformed into an immersive 3D installation that allows viewers to walk through the visual field generated by the Inception CNN and physically experience the mathematical distances between the images. The installation enables participants to walk randomly through the visual field, visit a specific cluster, or 823