Distribution Transformers represent priors as GMMs and use self- and cross-attention to produce posterior GMMs, achieving fast approximate Bayesian inference with on-the-fly prior adaptation.
Not Applicable (b) Descriptions of potential participant risks, with links to Institutional Review Board (IRB) approvals if applicable
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ML 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Distribution Transformers: Fast Approximate Bayesian Inference With On-The-Fly Prior Adaptation
Distribution Transformers represent priors as GMMs and use self- and cross-attention to produce posterior GMMs, achieving fast approximate Bayesian inference with on-the-fly prior adaptation.