Matt Buchholz

icon indicating a website link https://buchholzmd.github.io/ icon indicating an email [email protected]

icon company Courant Institute of Mathematical Sciences icon location New York, NY icon location span{ interests } = { mathematics of deep learning, bayesian deep learning, generative models, inverse problems, theory of computation } \n Math @ CIMS

Results 2 issues of Matt Buchholz

Best way to use scikit-learn distance functions for cost

I am trying to use the `sklearn.metrics.pairwise.haversine_distances` as the cost for SamplesLoss, in order to calculate the EMD for spherical distributions (see [https://arxiv.org/abs/2012.11116]([EMLight: Lighting Estimation via Spherical Distribution](https://arxiv.org/abs/2012.11116))). I am...

Regular softmax attention in last block

Hello, First, nice job on the work, I think this is a really interesting paper, with a lot of potential to enable further theoretical investigations into deep attention mechanisms. Looking...