Richard Duong, Viktor Stein, Robert Beinert, Johannes Hertrich, Gabriele Steidl: Wasserstein Gradient Flows of MMD Functionals with Distance Kernel and Cauchy Problems on Quantile Functions arxiv.org/abs/2408.07498 arxiv.org/pdf/2408.07498

On todays evening, I have the pleasure to give a talk with the topic of #Rényi regularized #optimaltransport at DWCAA24 in Canazei. A co-work with my colleague ViktorStein. #optimization #convex #optimal #transport

📝🚨 Finally we completed this one! How to compute the operator norm from only forward evaluations?

I'm at the wonderful CIRM again! Today I will present a poster about our work on #Wasserstein #gradientflows of MMD-#regularized of f-divergences at the SIGMA workshop jdigne.github.io/sigma2024/inde… organized by Albert Cohen, Julie Digne, Jalal Fadili, Olga Mula and Anthony Nouy.

It's been a wonderful SIGMA workshop at CIRM with great talks by Gabriel Peyré, Justin Solomon, Simon Boisserée, @sellan_s, Raphaël Barboni, Jean Feydy, Pierre Marion, Jonny and so many others. The second photo is the view from by bathroom window😁

PLS SHARE: I'm hiring a PhD student to work on ML theory, to begin in Fall 2025. Topics include: generalization bounds & statistical inference via online prediction, representation learning via optimal transport, sequential decision making... More info: cs.bme.hu/~gergo/jobs.ht…

I have multiple openings for M2 internship / PhD / postdoc in Nice (France) on topics related to bilevel optimization, automatic differentiation and safe machine learning. More details on my webpage samuelvaiter.com Contact me by email, and feel free to forward/RT :)

Wasserstein Gradient Flows for Moreau Envelopes of f-Divergences in Reproducing Kernel Hilbert Spaces ift.tt/6iaKofR

check this out: new postdoc program for AI-related research in Catalunya! our group is looking to hire within this program, ideally to work on topics related to RL theory. in case you're interested, pls DM or email me. (retweets appreciated!) ramonllull-aira.eu/application

How to compute || A - V ||, i.e. the adjoint mismatch of two linear maps, if we only have access to x -> Ax and y -> V*y. Together with Dirk L., we finally give the answer: arxiv.org/abs/2503.21361

Deadline: April 2nd, AOE 7th International Conference on "Geometric Science of Information" (GSI) Submission: cmt3.research.microsoft.com/GSI2025/ Web site: conference-gsi.org/gsi-2025/

I had a blast attending "the unpronounceable workshop" (according to Marie-Therese Wolfram) EYAWKADANAJKOS (Everything you always wanted to know about ... of the JKO scheme) at the University of Lyon 1! Thanks to Filippo and Anastasiia for organizing, the talks were truly wonderful!

There is no better blog (that I have found) which explains intuitively WHY momentum works.

Geometric calculations on probability manifolds from reciprocal relations in Master equations. arxiv.org/abs/2504.19368