Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore a 14-minute conference presentation from OOPSLA 2025 that introduces MultiPPL, a groundbreaking probabilistic multi-language framework enabling seamless interoperability between different probabilistic programming languages. Learn how researchers from Northeastern University address the current limitation where probabilistic programmers must commit to a single language upfront and cannot leverage the strengths of multiple specialized languages for heterogeneous programs. Discover the syntax and semantics of MultiPPL, which allows programmers to combine a high-performance exact discrete inference strategy with approximate importance sampling techniques within the same program. Examine the theoretical foundations including soundness proofs for the inference algorithm and review empirical evidence demonstrating how this approach enables complex heterogeneous probabilistic programming while simultaneously exploiting the complementary strengths and weaknesses of different probabilistic programming languages. Access the accompanying research article, supplementary materials with reproducible artifacts, and gain insights into multi-language semantics and Bayesian inference methodologies that could transform how probabilistic programming environments handle diverse computational requirements.
