Grouped Reverse Importance Sampling for the Partition Function

2026-06-25

AI Summary

This work addresses the challenge of efficiently estimating the partition function from samples drawn from a Boltzmann distribution. Traditional reverse importance sampling (RIS) suffers from high error due to its use of independent per-sample weights. To overcome this limitation, the authors propose Grouped Reverse Importance Sampling (GRIS), which partitions samples into groups of size $k \geq 2$ and introduces a joint weighting function that couples samples within each group to minimize mean squared error (MSE). Theoretically, they prove that the optimal weight depends solely on the total energy of the group, significantly simplifying both analysis and implementation, and establish—for the first time—that such coupled weights strictly outperform conventional RIS. Empirically, non-overlapping GRIS with $k=2,3$ reduces MSE by 20%–65% across three benchmark tasks, with sliding-window variants (FSW/VSW) yielding further improvements.

Abstract

We introduce and analyze several grouped variants of the method of reverse importance sampling (RIS) for estimating a partition function from samples of the Boltzmann distribution $p(x)=e^{ \betaU(x)}/Z(β)$. Ordinary RIS weighs each sample separately. By contrast, our proposed grouped RIS (GRIS) methods are based on assigning the samples into groups (or batches) of size $k\ge 2$ and applying a joint weight function to each group. The focal point of the research is the quest for a tractable weight function that would yield the smallest possible mean squared error (MSE). A simple identity relates the normalized MSE to the chi-squared divergence between the joint-weight distribution and the distribution of the $k$-fold sum of independent energies. Our first theoretical finding is that any weight that improves on ordinary RIS ($k=1$) must couple the group components. In other words, it must not be a product-form function across those components, as product-form weight functions always worsen the MSE. Our second, and more important, finding is that, without loss of optimality, it is sufficient to seek weight functions that depend only on the total energy, $\sum_iU(x_i)$, of the group (group-energy weight functions); for the sliding-window variants, the analogous result is open. This finding simplifies both the theoretical analysis and the application of the method substantially. For $k=2$ and $k=3$, the MSE associated with non-overlapping (NOL) groups is reduced by $20$--$65\%$ across three examples. We then propose two additional variants of GRIS, both based on sliding-window grouping (as opposed to NOL grouping). The first applies a fixed weight sliding window (FSW) across all (cyclic) shifts of the sliding window, and the second allows a variable-weight sliding window (VSW). The FSW scheme improves on the NOL one, and the VSW improves even further, as will be demonstrated numerically.

Problem

• partition function
• reverse importance sampling
• mean squared error
• Boltzmann distribution
• grouped sampling

Innovation

• grouped reverse importance sampling
• partition function estimation
• mean squared error
• chi-squared divergence
• sliding-window grouping