Dan Piponi on Nostr: nprofile1q…cfh5j Typically people doing this stuff are working in some space of ...

nprofile1qy2hwumn8ghj7un9d3shjtnyd968gmewwp6kyqpqgf8x0vypj2ve7hyuj8nxrspplavg56y5rc3j2zue0w2s9fphgj0s0cfh5j (nprofile…fh5j) Typically people doing this stuff are working in some space of parameters for a family of distributions where Riemannian metrics make sense. But I can imagine some people don't...