Sound Abstraction and Decomposition of Probabilistic Programs

Authors: Steven Holtzen, Guy Broeck, Todd Millstein

ICML 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimentally, we also illustrate the practical benefits of our framework as a tool to decompose probabilistic program inference.
Researcher Affiliation Academia 1University of California, Los Angeles. Correspondence to: Steven Holtzen <sholtzen@cs.ucla.edu>, Guy Van den Broeck <guyvdb@cs.ucla.edu>, Todd Millstein <todd@cs.ucla.edu>.
Pseudocode Yes Algorithm 1 Domain completion
Open Source Code No The paper mentions using third-party tools like 'Psi4 (Gehr et al., 2016)' and 'Web PPL (Goodman & Stuhlmüller, 2014)', but it does not provide an explicit statement or link for the open-source code of their own methodology.
Open Datasets No The paper describes programs used as examples for experiments (e.g., Multiplication, Markov Chain, Shuffle, and a program with noisy observations), but it does not refer to them as publicly available datasets nor does it provide concrete access information (links, DOIs, repositories, or formal citations) for them.
Dataset Splits No The paper describes various experimental setups but does not specify dataset split information (e.g., percentages, sample counts, or references to predefined splits) for training, validation, or testing.
Hardware Specification No The paper does not provide specific hardware details (e.g., CPU/GPU models, memory) used for running the experiments.
Software Dependencies Yes We utilized build 5334524fe.
Experiment Setup No The paper describes the probabilistic programs and inference methods, but it does not provide specific experimental setup details such as hyperparameters (e.g., learning rates, batch sizes, number of epochs) or specific optimizer settings for the MCMC experiments.