AI Paper Atlas

Research map / Optimization / Optimizers

Large-Scale Wasserstein Gradient Flows

Petr Mokrov, Alexander Korotin, Lingxiao Li, Aude Genevay, Justin Solomon, Evgeny Burnaev

NeurIPS 2021 · Optimization › Optimizers · approach: Stein variational gradient descent · 8 citations (OpenAlex)

Read the paper

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Use the ICNN-based Wasserstein gradient-flow sampler instead of SVGDBayesian logistic regression · germanaccuracy: 0.65 -> 0.67Improved
Use the ICNN-based Wasserstein gradient-flow sampler instead of SVGDBayesian logistic regression · bananaaccuracy: 0.54 -> 0.55Improved
Use the ICNN-based Wasserstein gradient-flow sampler instead of SVGDBayesian logistic regression · waveformaccuracy: 0.765 -> 0.78Improved
Use the ICNN-based Wasserstein gradient-flow sampler instead of SVGDBayesian logistic regression · imageaccuracy: 0.815 -> 0.82Improved
Use the ICNN-based Wasserstein gradient-flow sampler instead of SVGDBayesian logistic regression · diabetisaccuracy: 0.78 -> 0.775Hurt
Use the ICNN-based Wasserstein gradient-flow sampler instead of SVGDBayesian logistic regression · spliceaccuracy: 0.85 -> 0.845Hurt
Use the ICNN-based Wasserstein gradient-flow sampler instead of SVGDBayesian logistic regression · diabetislog-likelihood: -0.46 -> -0.45Improved
Use the ICNN-based Wasserstein gradient-flow sampler instead of SVGDBayesian logistic regression · twonormlog-likelihood: -0.062 -> -0.059Improved
Use the ICNN-based Wasserstein gradient-flow sampler instead of SVGDBayesian logistic regression · imagelog-likelihood: -0.44 -> -0.43Improved
Use the ICNN-based Wasserstein gradient-flow sampler instead of SVGDBayesian logistic regression · waveformlog-likelihood: -0.465 -> -0.485Hurt
Use the ICNN-based Wasserstein gradient-flow sampler instead of SVGDBayesian logistic regression · covtypeaccuracy: 0.75 -> 0.75No clear effect
Use the ICNN-based Wasserstein gradient-flow sampler instead of SVGDBayesian logistic regression · covtypelog-likelihood: -0.515 -> -0.515No clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy