Annealed Importance Sampling with q-Paths
This work provides an incremental improvement to the theoretical understanding and applicability of Annealed Importance Sampling for researchers working on partition function estimation.
This paper explores Annealed Importance Sampling (AIS) using q-paths, which generalize the standard geometric path. While AIS provides an unbiased estimator for partition functions, previous work mainly focused on geometric mixture paths; this work expands the applicability to q-paths.
Annealed importance sampling (AIS) is the gold standard for estimating partition functions or marginal likelihoods, corresponding to importance sampling over a path of distributions between a tractable base and an unnormalized target. While AIS yields an unbiased estimator for any path, existing literature has been primarily limited to the geometric mixture or moment-averaged paths associated with the exponential family and KL divergence. We explore AIS using $q$-paths, which include the geometric path as a special case and are related to the homogeneous power mean, deformed exponential family, and $α$-divergence.