SAIV 2025 SAIV 2025

The 8th International Symposium on AI Verification in Zagreb on July 21–22, 2025


Formal reasoning about learning systems presents novel, challenging, and exciting problems, including but not limited to the verification of neural networks and the development of safeguards for machine learning algorithms. The Symposium on AI Verification brings together researchers from the fields of formal methods and artificial intelligence, providing a platform for the exchange of ideas and cross-pollination of insights on these critical areas.

SAIV 2025 will be co-located with the 37th International Conference on Computer Aided Verification in Zagreb, Croatia.

SAIV 2025 will host the 6th International Verification of Neural Networks Competition (VNN-COMP’25).

Invited Speakers and Round Tables

Eleonora Giunchiglia Eleonora Giunchiglia (Imperial-X, London, UK)
Neuro-Symbolic and Generative AI
Guillermo Perez Guillermo A. Pérez (University of Antwerp, Belgium)
Decidable Problems for Partially Observable MDPs
Luca Arnaboldi Luca Arnaboldi (University of Birmingham, UK)
Formal Methods for Natural Language Processing
Andreea Costea Andreea Costea (TU Delft, The Netherlands)
Assured Automatic Programming via LLMs
Masaki Waga Masaki Waga (Kyoto University, Japan)
Soft Pattern Matching: Toward Runtime Verification of NLP Systems
Nora Ammann Nora Ammann (Advanced Research and Invention Agency, UK)
Safeguarded AI
The travel of our invited speakers was in part supported by The ELSA Mobility Program.