AI Paper Atlas

Research map / Probabilistic and scientific modeling / Variational and generative sampling

Neural Spline Flows

Conor Durkan, Artur Bekasov, Iain Murray, George Papamakarios

NeurIPS 2019 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Normalizing flow training and design · 27 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

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
Replace affine elementwise transforms with rational-quadratic splines in coupling flowsGlow-style coupling flow · GAStest log likelihood (nats): 12.24 ± 0.03 -> 13.09 ± 0.02Improved
Replace affine elementwise transforms with rational-quadratic splines in autoregressive flowsAutoregressive flow · HEPMASStest log likelihood (nats): -17.03 ± 0.02 -> -14.01 ± 0.03Improved
Replace affine transforms with rational-quadratic splines in coupling flowsGlow-style coupling flow · CIFAR-10 8-bitBPD: 3.41 -> 3.38Improved
Replace affine transforms with rational-quadratic splines in coupling flowsGlow-style coupling flow · ImageNet64 5-bitBPD: 1.81 -> 1.77Improved
Replace affine transforms with rational-quadratic splines in coupling flowsGlow-style coupling flow · ImageNet64 8-bitBPD: 3.91 -> 3.82Improved
Replace affine transforms with rational-quadratic splines in coupling flowsGlow-style coupling flow · CIFAR-10 5-bitBPD: 1.70 -> 1.70No clear effect
Use RQ-NSF instead of the standard-normal prior and diagonal-normal posteriorVAE · MNISTELBO (nats): -85.61 ± 0.51 -> -82.08 ± 0.46Improved
Use RQ-NSF instead of the standard-normal prior and diagonal-normal posteriorVAE · EMNISTELBO (nats): -125.89 ± 0.41 -> -119.49 ± 0.40Improved
Use RQ-NSF coupling flows rather than affine coupling flowsVAE · MNISTELBO (nats): -82.25 ± 0.46 -> -82.08 ± 0.46No clear effect
Use RQ-NSF autoregressive flows rather than affine autoregressive flowsVAE · EMNISTELBO (nats): -119.85 ± 0.40 -> -119.49 ± 0.40No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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