Generative Flow Networks (GFlowNets) have been introduced as a method to sample a diverse set of candidates in an active learning context, with a training objective that makes them approximately sample in proportion to a given reward function. In this paper, we show a number of additional theoretical properties of GFlowNets. They can be used to estimate joint probability distributions and the corresponding marginal distributions where some variables are unspecified and, of particular interest, can represent distributions over composite objects like sets and graphs. GFlowNets amortize the work typically done by computationally expensive MCMC methods in a single but trained generative pass. They could also be used to estimate partition functions and free energies, conditional probabilities of supersets (supergraphs) given a subset (subgraph), as well as marginal distributions over all supersets (supergraphs) of a given set (graph). We introduce variations enabling the estimation of entropy and mutual information, sampling from a Pareto frontier, connections to reward-maximizing policies, and extensions to stochastic environments, continuous actions and modular energy functions.

翻译：生成流网络（GFlowNets）最初被提出作为一种在主动学习背景下对多样化候选样本进行采样的方法，其训练目标使其能够近似地按照给定奖励函数的比例进行采样。本文展示了GFlowNets的若干附加理论特性。它们可用于估计联合概率分布及对应的边缘分布（当某些变量未指定时），尤其值得注意的是，它们能够表示复合对象（如集合和图）上的分布。GFlowNets通过单次训练后的生成过程，分摊了通常需要计算昂贵的MCMC方法完成的工作。它们还可用于估计配分函数与自由能、给定子集（子图）条件下超集（超图）的条件概率，以及给定集合（图）所有超集（超图）上的边缘分布。我们提出了若干变体方法，使其能够估计熵与互信息、从帕累托前沿采样、建立与奖励最大化策略的关联，并扩展至随机环境、连续动作和模块化能量函数场景。