Formal verification of neuro-symbolic cyber-physical systems, such as drones, medical devices and robots, is complicated. Neural components must be trained to be optimal with respect to the available data as well as the safety specifications, and then verified using specialised solvers. Symbolic models of the "cyber" and "physical" behaviour of the system must be constructed and verified in interactive theorem provers (ITPs), often requiring mature mathematical libraries to reason about the interplay of discrete and continuous dynamics, preferably obtaining infinite time-horizon guarantees. Finally, the results of the two already challenging verification tasks need to be integrated into a single proof in a coherent and consistent way, whilst preserving deployability of the resulting model. In this paper we present a compositional methodology for constructing such proofs. The Vehicle framework provides a functional, domain-specific language for specifying, training, and verifying neural components. We extend Vehicle to allow integration with any ITP with minimal effort, thereby bridging the gap between the neural and symbolic proofs. First, we describe how Vehicle’s standard bidirectional type checker can be reused to transpile neural specifications into an intermediate representation targeting multiple theorem provers. Second, we integrate Vehicle with Rocq, Isabelle/HOL, Agda and the industrial prover Imandra; and showcase a generic infinite time-horizon safety proof of a discrete cyber-physical system with a neural network controller in each ITP. Finally, to put the idea of compositional neural-cyber-physical system verification to the test, we use the Mathematical Components libraries in Rocq to verify infinite time-horizon safety of a medical device, modelled as a continuous cyber-physical system with a neural controller. To our knowledge, this is the first result of this kind in a general purpose ITP; and a result that was only feasible thanks to the compositionality provided by Vehicle's functional interface.

Concentration inequalities are standard lemmas providing upper bounds on deviations of random variables. To formalize concentration inequalities, we have been developing a general library of lemmas for probability theory in the Rocq prover. This effort led us to revisit already established technical aspects of the Mathematical Components libraries. In this paper, we report on improvements of general interest resulting from our formalization. We devise types for numeric values and a lightweight semi-decision procedure, based on interval arithmetic. We also extend the hierarchy of available mathematical structures to formalize Lebesgue spaces. We illustrate our new formalization of probability theory with the complete proof of a concentration inequality for Bernoulli sampling.

Abstract Focusing is a known technique for reducing the number of proofs while preserving derivability. Skolemisation is another technique designed to improve proof search, which reduces the number of back-tracking steps by representing dependencies on the term level and instantiate witness terms during unification at the axioms or fail with an occurs-check otherwise. Skolemisation for classical logic is well understood, but a practical skolemisation procedure for focused intuitionistic linear logic has been elusive so far. In this paper we present a focused variant of first-order intuitionistic linear logic together with a sound and complete skolemisation procedure.

For performance and verification in machine learning, new methods have recently been proposed that optimise learning systems to satisfy formally expressed logical properties. Among these methods, differentiable logics (DLs) are used to translate propositional or first-order formulae into loss functions deployed for optimisation in machine learning. At the same time, recent attempts to give programming language support for verification of neural networks showed that DLs can be used to compile verification properties to machine-learning backends. This situation is calling for stronger guarantees about the soundness of such compilers, the soundness and compositionality of DLs, and the differentiability and performance of the resulting loss functions. In this paper, we propose an approach to formalise existing DLs using the Mathematical Components library in the Coq proof assistant. Thanks to this formalisation, we are able to give uniform semantics to otherwise disparate DLs, give formal proofs to existing informal arguments, find errors in previous work, and provide formal proofs to missing conjectured properties. This work is meant as a stepping stone for the development of programming language support for verification of machine learning.

Recent work has shown that cell phone mobility data has the unique potential to create accurate models for human mobility and consequently the spread of infected diseases [74]. While prior studies have exclusively relied on a mobile network operator’s subscribers’ aggregated data in modelling disease dynamics, it may be preferable to contemplate aggregated mobility data of infected individuals only. Clearly, naively linking mobile phone data with health records would violate privacy by either allowing to track mobility patterns of infected individuals, leak information on who is infected, or both. This work aims to develop a solution that reports the aggregated mobile phone location data of infected individuals while still maintaining compliance with privacy expectations. To achieve privacy, we use homomorphic encryption, validation techniques derived from zero-knowledge proofs, and differential privacy. Our protocol’s open-source implementation can process eight million subscribers in 70 minutes.

The operation of trimming data sets is heavily used in AI systems. Trimming is useful to make AI systems more robust against adversarial or common perturbations. At the core of robust AI systems lies the concept that outliers in a data set occur with low probability, and therefore can be discarded with little loss of precision in the result. The statistical argument that formalizes this concept of robustness is based on an extension of the Chebyshev’s inequality first proposed by Tukey in 1960.

Communication protocols often rely on stateful mechanisms to ensure certain security properties. For example, counters and timestamps can be used to ensure authentication, or the security of communication can depend on whether a particular key is registered to a server or it has been revoked. ProVerif, like other state of the art tools for protocol analysis, achieves good performance by converting a formal protocol specification into a set of Horn clauses, that represent a monotonically growing set of facts that a Dolev-Yao attacker can derive from the system. Since this set of facts is not state-dependent, the category of protocols of our interest cannot be precisely analysed by such tools, as they would report false attacks due to the over-approximation. In this paper we present Set-π, an extension of the Applied π-calculus that includes primitives for handling databases of objects, and propose a translation from Set-π into Horn clauses that employs the set-membership abstraction to capture the non-monotonicity of the state. Furthermore, we give a characterisation of authentication properties in terms of the set properties in the language, and prove the correctness of our approach. Finally we showcase our method with three examples, a simple authentication protocol based on counters, a key registration protocol, and a model of the Yubikey security device.