Recent advances in AI and machine learning are reshaping the landscape of CS and mathematics research. For instance, systems like AlphaEvolve have discovered novel, more efficient algorithms for fundamental computations such as matrix multiplication, surpassing human-derived solutions. In the realm of formal mathematics, AI is being used to guide interactive theorem provers and even formalize complex proofs, bridging the gap between informal mathematical intuition and machine-checkable rigor. These developments suggest a future in which AI serves as a powerful collaborator for researchers, automating tedious tasks, uncovering novel insights, and expanding the frontier of knowledge.
This workshop will help bring the CAV community up to speed with these developments and foster a dialogue between AI/ML experts working on these topics and formal methods researchers. Specific objectives include:
Introducing the formal methods community to contemporary AI techniques for automated proof and discovery of algorithms, constructions, and counterexamples.
Discussing the unique challenges of applying AI to structured, logical domains.
Exploring the potential for neurosymbolic systems that combine the deductive power of traditional solvers with the pattern-recognition strengths of neural networks.
Identifying challenge problems that can benefit from emerging methods for AI-aided reasoning.
In addition to invited talks, the workshop will include a poster presentation session and several lightning talks. We will only accept submissions based on research that has already been publicly released in the last year (from January 2025 until now). There is no restriction on the venue where the work was published; it is acceptable for the work to have only been published as an Arxiv preprint. Our only requirements are that the work be high-quality and relevant to the themes of the workshop. Please present full information about where to find the work in the form below.
A two-person program committee (Swarat Chaudhuri and Armando Solar-Lezama) will identify the successful submissions. Authors of successful submissions will be invited to present posters, and in a subset of cases, lightning talks, at the event.
Please submit your proposals on or before May 31 at:
https://forms.gle/RESRT5uXXtZkVzjr9
Stanford University
University of California San Diego
Chalmers University
Brown University
Google DeepMind & UT Austin
Rice University
Miromind
ETH Zurich