Call for Papers

SAIV invites submissions presenting theoretical results, novel algorithms, tool development, and empirical evaluation.

Important Dates

All dates are 11:59AM (noon) UTC

• Abstract Registration: April 17
• Paper Submission: April 24
• Notification: May 21

Topics

The topics covered by SAIV include, but are not limited to, the following:

Formal methods for artificial intelligence

• Safety specifications for systems with learning components
• Symbolic analysis of cyber-physical systems with AI components
• Formal verification of neural networks
• Neuro-symbolic reasoning for AI safety
• Testing approaches for systems with AI components
• Formal guarantees for interpretable AI

Artificial intelligence for formal methods

• Machine Learning for program synthesis and control synthesis
• Machine Learning for automated reasoning and theorem-proving
• Statistical approaches to falsification and verification
• Data-driven verification
• Differentiable proof certificates

Submission Guidelines

→ OpenReview submission website

We invite four categories of submissions:

1. Original papers: describe original research and sufficient detail to assess the merits and relevance of the submission. For papers reporting experimental results, authors are strongly encouraged to make their data available. We welcome both short and long papers. There are no strict page limits, although we recommend not going below 6 pages and not exceeding 18 pages.
2. Presentation-only papers: describe work recently published or submitted. We see this as a way to provide additional access to important developments that the SAIV attendees may be unaware of.
3. Extended abstracts: describe preliminary work in progress. These reports may range in length from very short to full papers, and will be judged based on the expected level of interest for the community.
4. Benchmarks and Case Studies: showcasing practical evaluations, real-world applications, or lessons learned in verifying and deploying safety-critical AI systems, focusing on metrics, methodologies, and safety outcomes.​

SAIV 2025 uses a single-blind policy, so submissions need not be anonymized.

We require to use of the LNCS template. All papers conforming to the submission guidelines will be peer-reviewed by members of the program committee. Submissions will be evaluated on the basis of originality, importance of contribution, soundness, quality of presentation, and appropriate comparison to related work.

Papers selected for publications will appear in the SAIV 2025 conference proceedings in the LNCS series.