AI Paper Atlas

Research map / Optimization / Optimizers

Convex Potential Flows: Universal Probability Distributions with Optimal Transport and Convex Optimization

Chin-Wei Huang, Ricky T. Q. Chen, Christos Tsirigotis, Aaron Courville

ICLR 2021 · Optimization › Optimizers · approach: Stein variational gradient descent · 7 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 CP-Flow instead of the strongest listed baselinetabular density estimation · Powernegative log-likelihood (nats): -0.62$\,\pm\,$0.01 -> -0.52$\,\pm\,$0.01Hurt
Use CP-Flow instead of the strongest listed baselinetabular density estimation · Gasnegative log-likelihood (nats): -11.96$\,\pm\,$0.33 -> -10.36$\,\pm\,$0.03Hurt
Use CP-Flow instead of the strongest listed baselinetabular density estimation · HEPMASSnegative log-likelihood (nats): 14.92$\,\pm\,$0.08 -> 16.93$\,\pm\,$0.08Hurt
Use CP-Flow instead of the strongest listed baselinetabular density estimation · MINIBOONEnegative log-likelihood (nats): 8.86$\,\pm\,$0.15 -> 10.58$\,\pm\,$0.07Hurt
Use CP-Flow instead of the strongest listed baselinetabular density estimation · BSDS300negative log-likelihood (nats): -157.73$\,\pm\,$0.04 -> -154.99$\,\pm\,$0.08Hurt
Use CP-Flow instead of the strongest listed baselineimage density estimation · MNISTbits/dim: 0.97 -> 1.02Hurt
Use CP-Flow instead of the strongest listed baselineimage density estimation · CIFAR-10bits/dim: 3.28 -> 3.40Hurt
Replace CP-Flow blocks with coupling blocks (ablation)CP-Flow image model · MNISTbits/dim: 1.02 -> 1.02No clear effect
Replace CP-Flow blocks with coupling blocks (ablation)CP-Flow image model · CIFAR-10bits/dim: 3.40 -> 3.58Hurt
Replace CP-Flow blocks with invertible residual blocks (ablation)CP-Flow image model · MNISTbits/dim: 1.02 -> 1.04Hurt
Replace CP-Flow blocks with invertible residual blocks (ablation)CP-Flow image model · CIFAR-10bits/dim: 3.40 -> 3.46Hurt
Add input-augmented ICNN architecture (ablation)CP-Flow variational inference · Freyfacesnegative ELBO: 4.47 -> 4.45Improved

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