Matches in SemOpenAlex for { <https://semopenalex.org/work/W2989574682> ?p ?o ?g. }
Showing items 1 to 81 of
81
with 100 items per page.
- W2989574682 endingPage "8511" @default.
- W2989574682 startingPage "8499" @default.
- W2989574682 abstract "Energy-based models (EBMs) are powerful probabilistic models, but suffer from intractable sampling and density evaluation due to the partition function. As a result, inference in EBMs relies on approximate sampling algorithms, leading to a mismatch between the model and inference. Motivated by this, we consider the sampler-induced distribution as the model of interest and maximize the likelihood of this model. This yields a class of energy-inspired models (EIMs) that incorporate learned energy functions while still providing exact samples and tractable log-likelihood lower bounds. We describe and evaluate three instantiations of such models based on truncated rejection sampling, self-normalized importance sampling, and Hamiltonian importance sampling. These models out-perform or perform comparably to the recently proposed Learned Accept/RejectSampling algorithm and provide new insights on ranking Noise Contrastive Estimation and Contrastive Predictive Coding. Moreover, EIMs allow us to generalize a recent connection between multi-sample variational lower bounds and auxiliary variable variational inference. We show how recent variational bounds can be unified with EIMs as the variational family." @default.
- W2989574682 created "2019-12-05" @default.
- W2989574682 creator A5022202456 @default.
- W2989574682 creator A5048032272 @default.
- W2989574682 creator A5057996489 @default.
- W2989574682 creator A5086484914 @default.
- W2989574682 date "2019-10-31" @default.
- W2989574682 modified "2023-09-24" @default.
- W2989574682 title "Energy-Inspired Models: Learning with Sampler-Induced Distributions" @default.
- W2989574682 hasPublicationYear "2019" @default.
- W2989574682 type Work @default.
- W2989574682 sameAs 2989574682 @default.
- W2989574682 citedByCount "11" @default.
- W2989574682 countsByYear W29895746822019 @default.
- W2989574682 countsByYear W29895746822020 @default.
- W2989574682 countsByYear W29895746822021 @default.
- W2989574682 crossrefType "proceedings-article" @default.
- W2989574682 hasAuthorship W2989574682A5022202456 @default.
- W2989574682 hasAuthorship W2989574682A5048032272 @default.
- W2989574682 hasAuthorship W2989574682A5057996489 @default.
- W2989574682 hasAuthorship W2989574682A5086484914 @default.
- W2989574682 hasConcept C105795698 @default.
- W2989574682 hasConcept C106131492 @default.
- W2989574682 hasConcept C11413529 @default.
- W2989574682 hasConcept C126255220 @default.
- W2989574682 hasConcept C140779682 @default.
- W2989574682 hasConcept C154945302 @default.
- W2989574682 hasConcept C155846161 @default.
- W2989574682 hasConcept C19499675 @default.
- W2989574682 hasConcept C2776214188 @default.
- W2989574682 hasConcept C28826006 @default.
- W2989574682 hasConcept C31972630 @default.
- W2989574682 hasConcept C33923547 @default.
- W2989574682 hasConcept C41008148 @default.
- W2989574682 hasConcept C49937458 @default.
- W2989574682 hasConcept C52740198 @default.
- W2989574682 hasConceptScore W2989574682C105795698 @default.
- W2989574682 hasConceptScore W2989574682C106131492 @default.
- W2989574682 hasConceptScore W2989574682C11413529 @default.
- W2989574682 hasConceptScore W2989574682C126255220 @default.
- W2989574682 hasConceptScore W2989574682C140779682 @default.
- W2989574682 hasConceptScore W2989574682C154945302 @default.
- W2989574682 hasConceptScore W2989574682C155846161 @default.
- W2989574682 hasConceptScore W2989574682C19499675 @default.
- W2989574682 hasConceptScore W2989574682C2776214188 @default.
- W2989574682 hasConceptScore W2989574682C28826006 @default.
- W2989574682 hasConceptScore W2989574682C31972630 @default.
- W2989574682 hasConceptScore W2989574682C33923547 @default.
- W2989574682 hasConceptScore W2989574682C41008148 @default.
- W2989574682 hasConceptScore W2989574682C49937458 @default.
- W2989574682 hasConceptScore W2989574682C52740198 @default.
- W2989574682 hasLocation W29895746821 @default.
- W2989574682 hasOpenAccess W2989574682 @default.
- W2989574682 hasPrimaryLocation W29895746821 @default.
- W2989574682 hasRelatedWork W1959608418 @default.
- W2989574682 hasRelatedWork W2121693152 @default.
- W2989574682 hasRelatedWork W2159868298 @default.
- W2989574682 hasRelatedWork W2409550820 @default.
- W2989574682 hasRelatedWork W2554637102 @default.
- W2989574682 hasRelatedWork W2803106361 @default.
- W2989574682 hasRelatedWork W2893749619 @default.
- W2989574682 hasRelatedWork W2950701726 @default.
- W2989574682 hasRelatedWork W2962771722 @default.
- W2989574682 hasRelatedWork W2963981733 @default.
- W2989574682 hasRelatedWork W2970607325 @default.
- W2989574682 hasRelatedWork W2982492459 @default.
- W2989574682 hasRelatedWork W3037733149 @default.
- W2989574682 hasRelatedWork W3085775742 @default.
- W2989574682 hasRelatedWork W3093888319 @default.
- W2989574682 hasRelatedWork W3106068426 @default.
- W2989574682 hasRelatedWork W3111366808 @default.
- W2989574682 hasRelatedWork W3133733630 @default.
- W2989574682 hasRelatedWork W3139437910 @default.
- W2989574682 hasRelatedWork W3205586694 @default.
- W2989574682 hasVolume "32" @default.
- W2989574682 isParatext "false" @default.
- W2989574682 isRetracted "false" @default.
- W2989574682 magId "2989574682" @default.
- W2989574682 workType "article" @default.