AI Paper Atlas

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

Implicit Generation and Modeling with Energy Based Models

Yilun Du, Igor Mordatch

NeurIPS 2019 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Posterior sampling with diffusion guidance

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
ensemble sampling from last 10 model snapshots (ablation)EBM · CIFAR-10 UnconditionalFID: 40.58 -> 38.2Improved
ensemble sampling from last 10 model snapshots (ablation)EBM · CIFAR-10 UnconditionalInception score: 6.02 -> 6.78Improved
EBM instead of PixelIQNlikelihood generative models · CIFAR-10 UnconditionalFID: 49.46 -> 40.58Improved
EBM instead of SNGANGAN generator · CIFAR-10 UnconditionalFID: 21.7 vs 40.58Hurt
EBM instead of SNGAN at 128x128 (smaller net, not converged)ImageNet 128x128 generator · ImageNet 128x128 ConditionalFID: 27.62 vs 43.7Hurt
EBM instead of PixelIQNconditional generative models · ImageNet 32x32 ConditionalFID: 22.99 -> 14.31Improved
EBM instead of feedforward FC networktrajectory model · Robot hand, unconditionalaverage Frechet distance: 33.28 -> 5.96Improved
EBM instead of feedforward FC networktrajectory model · Robot hand, conditionalaverage Frechet distance: 19.75 -> 8.97Improved
EBM instead of LwFcontinual learning · Split MNISTaccuracy: 24.17 -> 64.99Improved
EBM instead of conditional VAE with same likelihood-classification schemegenerative continual learner · Split MNISTaccuracy: 40.04 -> 64.99Improved
EBM instead of PixelCNN++likelihood models · OOD detection averageAUROC: 0.47 -> 0.62Improved
run PGD attack incorporating samplingEBM classifier · CIFAR-10 adversarial robustnessrobustnessHurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 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