<ul><li>Soft Contrastive Variational Inference (SoftCVI) is introduced, allowing a family of variational objectives to be derived through a contrastive estimation framework.</li><li>SoftCVI reframes the inference task as a contrastive estimation problem and does not require positive or negative samples.</li><li>SoftCVI learns by sampling the variational distribution and computing ground truth soft classification labels from the unnormalized posterior itself.</li><li>Empirical investigation shows that SoftCVI can form stable and effective objectives for Bayesian inference tasks, frequently outperforming other variational approaches.</li></ul>

SoftCVI: Contrastive variational inference with self-generated soft labels

Discover more