Abstract
A control system consists of a plant component and a controller which periodically computes a control input for the plant. We consider systems where the controller is implemented by a feedforward neural network with ReLU activations. The reachability problem asks, given a set of initial states, whether a set of target states can be reached. We show that this problem is undecidable even for trivial plants and fixed-depth neural networks with three inputs and outputs. We also show that the problem becomes semi-decidable when the plant as well as the input and target sets are given by automata over infinite words.
Originalsprog | Engelsk |
---|---|
Titel | Bridging the Gap Between AI and Reality - 1st International Conference, AISoLA 2023, Selected Papers |
Redaktører | Bernhard Steffen |
Antal sider | 15 |
Forlag | Springer Science+Business Media |
Publikationsdato | 2025 |
Sider | 455-469 |
ISBN (Trykt) | 9783031737404 |
DOI | |
Status | Udgivet - 2025 |
Begivenhed | 1st International Symposium on Leveraging Applications of Formal Methods, AISoLA 2023 - Crete, Grækenland Varighed: 23 okt. 2023 → 28 okt. 2023 |
Konference
Konference | 1st International Symposium on Leveraging Applications of Formal Methods, AISoLA 2023 |
---|---|
Land/Område | Grækenland |
By | Crete |
Periode | 23/10/2023 → 28/10/2023 |
Navn | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
---|---|
Vol/bind | 14129 LNCS |
ISSN | 0302-9743 |
Bibliografisk note
Publisher Copyright:© The Author(s) 2025.