Control-Data Separation and Logical Condition Propagation for Efficient Inference on Probabilistic Programs
This work addresses the efficiency problem in probabilistic program inference, benefiting users working with complex programs and rare events.
This paper introduces a new sampling framework for probabilistic programs, combining control-data separation and logical condition propagation. The approach significantly improves efficiency, particularly for programs with while loops and rare observations.
We present a novel sampling framework for probabilistic programs. The framework combines two recent ideas -- \emph{control-data separation} and \emph{logical condition propagation} -- in a nontrivial manner so that the two ideas boost the benefits of each other. We implemented our algorithm on top of Anglican. The experimental results demonstrate our algorithm's efficiency, especially for programs with while loops and rare observations.