Automatic Fault Detection for Deep Learning Programs Using Graph Transformations

Nowadays, we are witnessing an increasing demand in both corporates and academia for exploiting Deep Learning ( DL ) to solve complex real-world problems. A DL program encodes the network structure of a desirable DL model and the process by which the model learns from the training dataset. Like any software, a DL program can be faulty, which implies substantial challenges of software quality assurance, especially in safety-critical domains. It is therefore crucial to equip DL development teams with efficient fault detection techniques and tools. In this article, we propose NeuraLint , a model-based fault detection approach for DL programs, using meta-modeling and graph transformations. First, we design a meta-model for DL programs that includes their base skeleton and fundamental properties. Then, we construct a graph-based verification process that covers 23 rules defined on top of the meta-model and implemented as graph transformations to detect faults and design inefficiencies in the generated models (i.e., instances of the meta-model). First, the proposed approach is evaluated by finding faults and design inefficiencies in 28 synthesized examples built from common problems reported in the literature. Then NeuraLint successfully finds 64 faults and design inefficiencies in 34 real-world DL programs extracted from Stack Overflow posts and GitHub repositories. The results show that NeuraLint effectively detects faults and design issues in both synthesized and real-world examples with a recall of 70.5% and a precision of 100%. Although the proposed meta-model is designed for feedforward neural networks, it can be extended to support other neural network architectures such as recurrent neural networks. Researchers can also expand our set of verification rules to cover more types of issues in DL programs.

Connectivity graphs: a method for proving deadlock freedom based on separation logic

We introduce the notion of a connectivity graph —an abstract representation of the topology of concurrently interacting entities, which allows us to encapsulate generic principles of reasoning about deadlock freedom . Connectivity graphs are parametric in their vertices (representing entities like threads and channels) and their edges (representing references between entities) with labels (representing interaction protocols). We prove deadlock and memory leak freedom in the style of progress and preservation and use separation logic as a meta theoretic tool to treat connectivity graph edges and labels substructurally. To prove preservation locally, we distill generic separation logic rules for local graph transformations that preserve acyclicity of the connectivity graph. To prove global progress locally, we introduce a waiting induction principle for acyclic connectivity graphs. We mechanize our results in Coq, and instantiate our method with a higher-order binary session-typed language to obtain the first mechanized proof of deadlock and leak freedom.

Cactus Graphs with Maximal Multiplicative Sum Zagreb Index

A connected graph G is said to be a cactus if any two cycles have at most one vertex in common. The multiplicative sum Zagreb index of a graph G is the product of the sum of the degrees of adjacent vertices in G. In this paper, we introduce several graph transformations that are useful tools for the study of the extremal properties of the multiplicative sum Zagreb index. Using these transformations and symmetric structural representations of some cactus graphs, we determine the graphs having maximal multiplicative sum Zagreb index for cactus graphs with the prescribed number of pendant vertices (cut edges). Furthermore, the graphs with maximal multiplicative sum Zagreb index are characterized among all cactus graphs of the given order.

Weisfeiler-Leman in the Bamboo: Novel AMR Graph Metrics and a Benchmark for AMR Graph Similarity

Several metrics have been proposed for assessing the similarity of (abstract) meaning representations (AMRs), but little is known about how they relate to human similarity ratings. Moreover, the current metrics have complementary strengths and weaknesses: Some emphasize speed, while others make the alignment of graph structures explicit, at the price of a costly alignment step. In this work we propose new Weisfeiler-Leman AMR similarity metrics that unify the strengths of previous metrics, while mitigating their weaknesses. Specifically, our new metrics are able to match contextualized substructures and induce n:m alignments between their nodes. Furthermore, we introduce a Benchmark for AMR Metrics based on Overt Objectives (Bamboo), the first benchmark to support empirical assessment of graph-based MR similarity metrics. Bamboo maximizes the interpretability of results by defining multiple overt objectives that range from sentence similarity objectives to stress tests that probe a metric’s robustness against meaning-altering and meaning- preserving graph transformations. We show the benefits of Bamboo by profiling previous metrics and our own metrics. Results indicate that our novel metrics may serve as a strong baseline for future work.