AI Paper Atlas

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

The continuous categorical: a novel simplex-valued exponential family

Elliott Gordon-Rodriguez, Gabriel Loaiza-Ganem, John Cunningham

ICML 2020 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Tractable probabilistic models · 11 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
Use the CC likelihood instead of the Dirichlet likelihoodLinear regression model · UK election data test setMAE: 0.105 -> 0.075Improved
Use the CC likelihood instead of the Dirichlet likelihoodLinear regression model · UK election data test setRMSE: 0.132 -> 0.101Improved
Use the CC likelihood instead of the Dirichlet likelihoodMLP regression model · UK election data test setMAE: 0.087 -> 0.072Improved
Use the CC likelihood instead of the Dirichlet likelihoodMLP regression model · UK election data test setRMSE: 0.112 -> 0.097Improved
Use the CC objective instead of Soft XEStudent network · MNIST test setaccuracy: 94.9% -> 95.6%Improved
Use the CC objective instead of Soft XEStudent network · MNIST test setRMSE: 0.029 -> 0.024Improved
Use teacher soft targets with Soft XE instead of original hard labelsStudent network · MNIST test setaccuracy: 93.2% -> 94.9%Improved
Increase the temperature of teacher soft targets (ablation)Dirichlet student model · MNISTtest errorImproved
Add L2 regularization to the regression network (ablation)Dirichlet UK election regression model · UK election datatest errorMixed
Replace vanilla LDA with a CC-prior topic modelTopic model · NeurIPS papers held-outperplexity per word: 557 -> 562Hurt
Fit a CC mean estimator instead of a Dirichlet mean estimatorSimplex-valued data estimator · Simulation studyempirical biasImproved
Use the CC objective instead of the Dirichlet likelihoodStudent network · MNIST test setaccuracy: 90.6% -> 95.6%Improved

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