A map of the latest 11 million papers split by semantic similarity and time slices [P]

I have building alternative ways explore scientifc literature. The goal was to make the large number of papers published daily easier to keep up with by visualising the macro scopic trend. It is free to use at The Global Research Space for any one interested in giving it a try! How I built it I sourced the latest 11M papers from OpenAlex and Arxiv and ecoded them using SPECTER 2 on titles and abstracts then projecting it down to 2d using UMAP and creating labels within voronoi bounds around high density pea