NCE, reverse logistic regression, multiple importance sampling, and bridge sampling are shown to be equivalent under specific conditions for energy-based models, yielding a unified estimator family.
Optimizing the noise in self-supervised learning: From importance sampling to noise-contrastive estimation
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CE 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
A unifying view of contrastive learning, importance sampling, and bridge sampling for energy-based models
NCE, reverse logistic regression, multiple importance sampling, and bridge sampling are shown to be equivalent under specific conditions for energy-based models, yielding a unified estimator family.