Abstract:In this paper we survey our recent results characterizing various graph neural network (GNN) architectures in terms of their ability to classify nodes over graphs, for classifiers based on unary logical formulas- or queries. We focus on the language FOC2, a well-studied fragment of FO. This choice is motivated by the fact that FOC2 is related to theWeisfeiler-Lehman (WL) test for checking graph isomorphism, which has the same ability as GNNs for distinguishing nodes on graphs. We unveil the exact relationship between FOC2 and GNNs in terms of node classification. To tackle this problem, we start by studying a popular basic class of GNNs, which we call AC-GNNs, in which the features of each node in a graph are updated, in successive layers, according only to the features of its neighbors. We prove that the unary FOC2 formulas that can be captured by an AC-GNN are exactly those that can be expressed in its guarded fragment, which in turn corresponds to graded modal logic. This result implies in particular that ACGNNs are too weak to capture all FOC2 formulas. We then seek for what needs to be added to AC-GNNs for capturing all FOC2. We show that it suffices to add readouts layers, which allow updating the node features not only in terms of its neighbors, but also in terms of a global attribute vector. We call GNNs with readouts ACR-GNNs. We also describe experiments that validate our findings by showing that, on synthetic data conforming to FOC2 but not to graded modal logic, AC-GNNs struggle to fit in while ACR-GNNs can generalise even to graphs of sizes not seen during training.

Graph Neural Networks and Boolean Satisfiability

Graph Neural Reasoning May Fail in Certifying Boolean Unsatisfiability

Graph Neural Networks for Propositional Model Counting

Graph Neural Networks for Reasoning 2 Quantified Boolean Formulas

Understanding GNNs for Boolean Satisfiability through Approximation Algorithms

Graph Representations for Higher-Order Logic and Theorem Proving

Graph Neural Reasoning for 2-Quantified Boolean Formula Solvers

The Logical Expressiveness of Graph Neural Networks

G4SATBench: Benchmarking and Advancing SAT Solving with Graph Neural Networks

GraSS: Combining Graph Neural Networks with Expert Knowledge for SAT Solver Selection

Graph Neural Networks for Maximum Constraint Satisfaction

Solving the Satisfiability Problem Through Boolean Networks

Using deep learning to construct stochastic local search SAT solvers with performance bounds

IB-Net: Initial Branch Network for Variable Decision in Boolean Satisfiability

Decidability of Graph Neural Networks via Logical Characterizations

Graph Reasoning Networks

Boolean-aware Boolean Circuit Classification: A Comprehensive Study on Graph Neural Network

LinSATNet: The Positive Linear Satisfiability Neural Networks

Can $Q$-Learning with Graph Networks Learn a Generalizable Branching Heuristic for a SAT Solver?

The Expressive Power of Graph Neural Networks as a Query Language

Graph Neural Networks for Binary Programming