AI Paper Atlas

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

VAEL: Bridging Variational Autoencoders and Probabilistic Logic Programming

Eleonora Misino, Giuseppe Marra, Emanuele Sansone

NeurIPS 2022 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Structured latent VAEs · 4 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 neural label head with ProbLog logic program prior (VAEL vs CCVAE)conditional VAE · 2digit MNISTpredictive accuracy: 0.5284 -> 0.8477Improved
replace neural label head with ProbLog logic program prior (VAEL vs CCVAE)conditional VAE · Mariopredictive accuracy: 1.0 -> 0.977No clear effect
replace neural label head with ProbLog logic program prior (VAEL vs CCVAE)conditional VAE · 2digit MNISTgenerative accuracy: 0.5143 -> 0.7922Improved
replace neural label head with ProbLog logic program prior (VAEL vs CCVAE)conditional VAE · Mariogenerative accuracy: 0.0 -> 0.8135Improved
replace neural label head with ProbLog logic program prior (VAEL vs CCVAE)conditional VAE · 2digit MNISTreconstruction loss: 1549 -> 1542No clear effect
replace neural label head with ProbLog logic program prior (VAEL vs CCVAE)conditional VAE · Marioreconstruction loss: 43461 -> 42734Improved
train VAEL with 10 images per pair vs CCVAE with 100 per pairVAEL · 2digit MNIST data efficiencydiscriminative accuracy: 0.121 -> 0.445Improved
train VAEL with 10 images per pair vs CCVAE with 100 per pairVAEL · 2digit MNIST data efficiencygenerative accuracy: 0.284 -> 0.415Improved
swap ProbLog program at test time to new tasks without retrainingVAEL · 2digit MNIST and Mario novel tasksqualitative generation consistencyImproved
more training images per pair for CCVAE on simplified 0-2 digit datasetCCVAE · simplified 2digit MNISTaccuracy: 2500 images per pairImproved

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