一句话：用行为类型系统驱动类型导向翻译，自动生成概率程序的 guide，保证可靠性并提升摊销推断效果。
在概率编程语言（PPL）中，基于优化的推断方法需要为给定模型程序构造可训练的 guide 程序，而推断的可靠性与有效性高度依赖 guide 的质量；通用 PPL 的强大表达力使这一构造极具挑战。本文提出一种在通用 PPL 中自动生成 guide 以支持深度摊销推断的方法：guide 通过一种新型行为类型系统驱动的类型导向翻译生成，并以语法方式提取和利用条件独立结构（其语义解释留待后续工作）。尽管通用 PPL 允许复杂控制流，所生成的 guide 仍被保证满足一项关键的可靠性条件；同时在一组基准上与当前最优基线相比，持续改善训练与推断表现。
In probabilistic programming languages (PPLs), a critical step in optimization-based inference methods is constructing, for a given model program, a trainable guide program. Soundness and effectiveness of inference rely on constructing good guides, but the expressive power of a universal PPL poses challenges. This paper introduces an approach to automatically generating guides for deep amortized inference in a universal PPL. Guides are generated using a type-directed translation per a novel behavioral type system. Guide generation extracts and exploits independence structures using a syntactic approach to conditional independence, with a semantic account left to further work. Despite the control-flow expressiveness allowed by the universal PPL, generated guides are guaranteed to satisfy a critical soundness condition and moreover, consistently improve training and inference over state-of-the-art baselines for a suite of benchmarks.
@article{paperbot2119,
title = {Type-Preserving, Dependence-Aware Guide Generation for Sound, Effective Amortized Probabilistic Inference},
author = {Jianlin Li and Leni Aniva and Pengyuan Shi and Yizhou Zhang},
journal = {Proceedings of the ACM on Programming Languages},
volume = {7},
number = {POPL},
year = {2023},
doi = {10.1145/3571243}
}