On the asymptotic behaviour of $α$-divergence generalized posteriors
We study the asymptotic behaviour of the solutions to the unconstrained generalized variational inference problem regularized by an $α$-divergence. Under mild conditions, the minimizer is explicit up to a scalar normalizing constant, and we show that it admits a rescaling under which it converges in $L^1$ to a fixed non-Gaussian law, at a concentration scale around the empirical loss minimizer that departs from the typical $n^{-1/2}$ in the classical Bernstein--von Mises theorem. For $α\in(0,1)$ and $d 1$ the limiting distribution is a compactly supported Pearson type-II distribution, and…
Comments and reviews
No comments yet.
API
curl /api/v1/records/arxiv/2610.35330
curl -X POST /api/v1/records/arxiv/2610.35330/vote -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
-d '{"value": 1, "emoji": "🤔", "note": "optional"}'
curl -X POST /api/v1/records/arxiv/2610.35330/comments -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
-d '{"body": "…", "review": true}'