Call for papers: 2nd Workshop on Machine Learning for Solvers and Provers (at FLoC 2026)

===========================================

The 2nd Workshop on Machine Learning for Solvers and Provers (ML4SP)

    July 18, 2026, Lisbon, Portugal

===========================================

Workshop page: https://ml4sp.github.io/

We warmly invite you to submit your work to the 2nd Workshop on Machine Learning for Solvers and Provers (ML4SP), organised as part of the 2026 Federated Logic Conference (FLoC 2026) at Lisbon, Portugal. The workshop will run during the first block of the conference, on July 18, 2026.

Topics

Machine learning (ML) has had a substantial impact on SAT/SMT and CP solvers, as well as automated theorem provers. Recent advances have demonstrated the power of ML to inform solver heuristics, guide proof search, and optimize algorithm portfolios. Despite growing interest in this direction, work on ML for solvers and provers is often scattered across multiple research communities – SAT, SMT, CP, theorem proving, formal methods, and machine learning – with few opportunities for focused interaction.

The ML4SP workshop aims to bring together researchers and practitioners working at the intersection of machine learning and formal reasoning systems. It provides a forum for the presentation of recent work, the exchange of ideas, and the fostering of collaboration between these communities.

Topics of interest include, but are not limited to, ML-driven approaches for:

Heuristics (branching, restarts, …) in CP, SAT, SMT, and MIP solvers
Tactic selection and proof guidance in automated and interactive theorem provers
Algorithm selection, parameter tuning and algorithm configuration, and portfolio solvers
End-to-end learning for solvers and provers
Benchmark generation and instance hardness prediction
Applications of ML-enhanced reasoning in verification, synthesis, planning, and related areas
Leveraging large language models (LLMs) for solver heuristics and proof guidance

Submission

We welcome submissions describing previously published work, ongoing research, and position papers and early-stage ideas intended to stimulate discussion.

Submission should be in PDF form, following the LIPIcs guidelines. They can be: