Off-policy learning (OPL) aims at finding improved policies from logged bandit data, often by
minimizing the inverse propensity scoring (IPS) estimator of the risk. In this work, we …

Adaptive importance sampling (AIS) algorithms are widely used to approximate expectations
with respect to complicated target probability distributions. When the target has heavy tails …

We introduce a new mean-field ODE and corresponding interacting particle systems for
sampling from an unnormalized target density or Bayesian posterior. The interacting particle …

This paper explores the connections between tempering (for Sequential Monte Carlo; SMC)
and entropic mirror descent to sample from a target probability distribution whose …

Inferring parameters of a latent variable model can be a daunting task when the conditional
distribution of the latent variables given the observed ones is intractable. Variational …

Driven by several successful applications such as in stochastic gradient descent or in
Bayesian computation, control variates have become a major tool for Monte Carlo …

We study the variational inference problem of minimizing a regularized R\'enyi divergence
over an exponential family, and propose a relaxed moment-matching algorithm, which …

We present {\em generative clustering}(GC) for clustering a set of documents, $\mathrm {X}
$, by using texts $\mathrm {Y} $ generated by large language models (LLMs) instead of by …

This paper addresses the problem of approximating an unknown probability distribution with
density $ f $--which can only be evaluated up to an unknown scaling factor--with the help of …

Monte Carlo methods have their roots in the 1940s and are closely associated with
mathematician Stanisław Ulam. While recovering from an illness, Ulam was playing solitaire …