From Well Structured Transition Systems to Program Verification
We describe the use of the theory of WSTS for verifying programs.
Source author record
Alain Finkel
Alain Finkel appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.
ResearcherUnclaimed source record
Catalog footprint
What is connected
7works
5topics
4close collaborators
Research graph
See the researcher in context
Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.
Building this map preview
BZPEER is loading the nearby papers, people, topics and institutions for this page.
Published work
7 published item(s)
preprint2020arXiv
Property-Directed Verification of Recurrent Neural Networks
This paper presents a property-directed approach to verifying recurrent neural networks (RNNs). To this end, we learn a deterministic finite automaton as a surrogate model from a given RNN using active automata learning. This model may then be analyzed using model checking as verification technique. The term property-directed reflects the idea that our procedure is guided and controlled by the given property rather than performing the two steps separately. We show that this not only allows us to discover small counterexamples fast, but also to generalize them by pumping towards faulty flows hinting at the underlying error in the RNN.
0 citations0 reviewsNo code linkedOpen access
preprint2016arXiv
Forward Analysis and Model Checking for Trace Bounded WSTS
We investigate a subclass of well-structured transition systems (WSTS), the bounded---in the sense of Ginsburg and Spanier (Trans. AMS 1964)---complete deterministic ones, which we claim provide an adequate basis for the study of forward analyses as developed by Finkel and Goubault-Larrecq (Logic. Meth. Comput. Sci. 2012). Indeed, we prove that, unlike other conditions considered previously for the termination of forward analysis, boundedness is decidable. Boundedness turns out to be a valuable restriction for WSTS verification, as we show that it further allows to decide all $ω$-regular properties on the set of infinite traces of the system.
0 citations0 reviewsNo code linkedOpen access
preprint2012arXiv
Forward Analysis for WSTS, Part II: Complete WSTS
We describe a simple, conceptual forward analysis procedure for infinity-complete WSTS S. This computes the so-called clover of a state. When S is the completion of a WSTS X, the clover in S is a finite description of the downward closure of the reachability set. We show that such completions are infinity-complete exactly when X is an omega-2-WSTS, a new robust class of WSTS. We show that our procedure terminates in more cases than the generalized Karp-Miller procedure on extensions of Petri nets and on lossy channel systems. We characterize the WSTS where our procedure terminates as those that are clover-flattable. Finally, we apply this to well-structured counter systems.
0 citations0 reviewsNo code linkedOpen access
preprint2012arXiv
Model Checking Vector Addition Systems with one zero-test
We design a variation of the Karp-Miller algorithm to compute, in a forward manner, a finite representation of the cover (i.e., the downward closure of the reachability set) of a vector addition system with one zero-test. This algorithm yields decision procedures for several problems for these systems, open until now, such as place-boundedness or LTL model-checking. The proof techniques to handle the zero-test are based on two new notions of cover: the refined and the filtered cover. The refined cover is a hybrid between the reachability set and the classical cover. It inherits properties of the reachability set: equality of two refined covers is undecidable, even for usual Vector Addition Systems (with no zero-test), but the refined cover of a Vector Addition System is a recursive set. The second notion of cover, called the filtered cover, is the central tool of our algorithms. It inherits properties of the classical cover, and in particular, one can effectively compute a finite representation of this set, even for Vector Addition Systems with one zero-test.
0 citations0 reviewsNo code linkedOpen access
preprint2011arXiv
Bounded Parikh Automata
The Parikh finite word automaton model (PA) was introduced and studied by Klaedtke and Ruess in 2003. Here, by means of related models, it is shown that the bounded languages recognized by PA are the same as those recognized by deterministic PA. Moreover, this class of languages is the class of bounded languages whose set of iterations is semilinear.
0 citations0 reviewsNo code linkedOpen access
preprint2011arXiv
On the expressiveness of Parikh automata and related models
The Parikh finite word automaton (PA) was introduced and studied by Klaedtke and Ruess in 2003. Natural variants of the PA arise from viewing a PA equivalently as an automaton that keeps a count of its transitions and semilinearly constrains their numbers. Here we adopt this view and define the affine PA (APA), that extends the PA by having each transition induce an affine transformation on the PA registers, and the PA on letters (LPA), that restricts the PA by forcing any two transitions on same letter to affect the registers equally. Then we report on the expressiveness, closure, and decidability properties of such PA variants. We note that deterministic PA are strictly weaker than deterministic reversal-bounded counter machines. We develop pumping-style lemmas and identify an explicit PA language recognized by no deterministic PA.
0 citations0 reviewsNo code linkedOpen access