Verified explanations are a principled way to explain the decisions taken by neural networks, which are otherwise black-box in nature. However, these techniques face significant scalability challenges, as they require multiple calls to neural network verifiers, each of them with an exponential worst-case complexity. We present FaVeX, a novel algorithm to compute verified explanations. FaVeX accelerates the computation by dynamically combining batch and sequential processing of input features, and by reusing information from previous queries, both when proving invariances with respect to certain input features, and when searching for feature assignments altering the prediction. Furthermore, we present a novel and hierarchical definition of verified explanations, termed verifier-optimal robust explanations, that explicitly factors the incompleteness of network verifiers within the explanation. Our comprehensive experimental evaluation demonstrates the superior scalability of both FaVeX, and of verifier-optimal robust explanations, which together can produce meaningful formal explanation on networks with hundreds of thousands of non-linear activations.

Due to its interdisciplinary nature, the development of data science software is particularly prone to a wide range of potential mistakes that can easily and silently compromise the final results. Several tools have been proposed that can help the data scientist in identifying the most common, low-level programming issues. However, these tools often fall short in detecting higher-level, domain-specific issues typical of data science pipelines, where subtle errors may not trigger exceptions but can still lead to incorrect or misleading outcomes, or unexpected behaviors. In this paper, we present PYRA , a static analysis tool that aims at detecting code smells in data science workflows. PYRA builds upon the Abstract Interpretation framework to infer abstract datatypes, and exploits such information to flag 16 categories of potential code smells concerning misleading visualizations, challenges for reproducibility, as well as misleading, unreliable or unexpected results. Unlike traditional linters, which focus on syntactic or stylistic issues, PYRA reasons over a domain-specific type system to identify data science-specific problems – such as improper data preprocessing steps and procedures’ misapplications – that could silently propagate through a data-manipulation pipeline. Beyond static checking, we envision tools like PYRA becoming integral components of the development loop, with analysis reports guiding correction and helping assess the reliability of machine learning pipelines. We evaluate PYRA on a benchmark suite of real-world Jupyter notebooks, showing its effectiveness in detecting practical data science issues, thereby enhancing transparency, correctness, and reproducibility in data science software.

Ethereum enables the creation and execution of decentralized applications through smart contracts, that are compiled to Ethereum Virtual Machine (EVM) bytecode. Once deployed in the blockchain, the bytecode is immutable; hence, ensuring that smart contracts are bug-free before their deployment is of utmost importance. A crucial preliminary step for any effective static analysis of EVM bytecode is the extraction of the control-flow graph (CFG): this presents significant challenges due to potentially statically unknown jump destinations. In this paper we present a novel approach, based on abstract interpretation, aiming at building a sound CFG from EVM bytecode smart contracts. Our analysis, which is implemented in our static analyzer EVMLiSA, is based on a parametric abstract domain that approximates concrete execution stacks at each program point as an l-sized set of abstract stacks of maximal height h; the results of the analysis are then used to resolve the jump destinations at jump nodes. In our preliminary experiments, by fine-tuning the analysis parameters, EVMLiSA builds sound CFGs for all smart contracts where permanent storage-related opcodes do not influence jump destinations.

Abstract In the context of abstract interpretation-based static analysis, we propose a new abstract operator modeling the split of control flow paths: the goal of the operator is to enable a more efficient analysis when using abstract domains that are computationally expensive, having no negative effect on precision, and occasionally resulting in a more precise analysis. We focus on the case of conditional branches guarded by numeric linear constraints, including implicit numerical branches. We provide an experimental evaluation of real-world test cases, showing that by using the split operator we can achieve significant efficiency improvements with respect to the classical approach for a static analysis based on the domain of convex polyhedra. We also briefly discuss the applicability of this new operator to different, possibly non-numeric abstract domains.