Semantic probabilistic layers for neuro-symbolic learning

K Ahmed, S Teso, KW Chang… - Advances in …, 2022 - proceedings.neurips.cc
We design a predictive layer for structured-output prediction (SOP) that can be plugged into
any neural network guaranteeing its predictions are consistent with a set of predefined …

Einsum networks: Fast and scalable learning of tractable probabilistic circuits

R Peharz, S Lang, A Vergari… - International …, 2020 - proceedings.mlr.press
Probabilistic circuits (PCs) are a promising avenue for probabilistic modeling, as they permit
a wide range of exact and efficient inference routines. Recent “deep-learning-style” …

Continuous mixtures of tractable probabilistic models

AHC Correia, G Gala, E Quaeghebeur… - Proceedings of the …, 2023 - ojs.aaai.org
Probabilistic models based on continuous latent spaces, such as variational autoencoders,
can be understood as uncountable mixture models where components depend continuously …

Understanding the distillation process from deep generative models to tractable probabilistic circuits

X Liu, A Liu, G Van den Broeck… - … Conference on Machine …, 2023 - proceedings.mlr.press
Abstract Probabilistic Circuits (PCs) are a general and unified computational framework for
tractable probabilistic models that support efficient computation of various inference tasks …

A survey of sum–product networks structural learning

R **a, Y Zhang, X Liu, B Yang - Neural Networks, 2023 - Elsevier
Sum–product networks (SPNs) in deep probabilistic models have made great progress in
computer vision, robotics, neuro-symbolic artificial intelligence, natural language …

Sum-product networks: A survey

R Sánchez-Cauce, I París… - IEEE Transactions on …, 2021 - ieeexplore.ieee.org
A sum-product network (SPN) is a probabilistic model, based on a rooted acyclic directed
graph, in which terminal nodes represent probability distributions and non-terminal nodes …

Sparse probabilistic circuits via pruning and growing

M Dang, A Liu… - Advances in Neural …, 2022 - proceedings.neurips.cc
Probabilistic circuits (PCs) are a tractable representation of probability distributions allowing
for exact and efficient computation of likelihoods and marginals. There has been significant …

Interventional sum-product networks: Causal inference with tractable probabilistic models

M Zečević, D Dhami, A Karanam… - Advances in neural …, 2021 - proceedings.neurips.cc
While probabilistic models are an important tool for studying causality, doing so suffers from
the intractability of inference. As a step towards tractable causal models, we consider the …

Group fairness by probabilistic modeling with latent fair decisions

YJ Choi, M Dang, G Van den Broeck - Proceedings of the AAAI …, 2021 - ojs.aaai.org
Abstract Machine learning systems are increasingly being used to make impactful decisions
such as loan applications and criminal justice risk assessments, and as such, ensuring …

Scaling up probabilistic circuits by latent variable distillation

A Liu, H Zhang, GV Broeck - arxiv preprint arxiv:2210.04398, 2022 - arxiv.org
Probabilistic Circuits (PCs) are a unified framework for tractable probabilistic models that
support efficient computation of various probabilistic queries (eg, marginal probabilities) …