Satisfiability Modulo Theories
Also known as SMT
Overview
SMT combines Boolean reasoning with specialized theory solvers to decide whether constraints can hold together, supporting applications such as software and neural-network verification, scheduling, and synthesis. A central approach is CDCL(T), which interleaves clause learning with theory checks; other methods use bit-blasting, optimization, or translations between SMT and constraint-programming or SAT solvers. Current work also explores proof-producing solvers and neuro-symbolic systems that use language models to propose constraints or lemmas which solvers can validate. A key tension is scaling to large or nonlinear problems while preserving exact, checkable guarantees.
Latest papers
23 papers1 paper in the last four weeks, with none the four weeks before. 0.0% of all new papers.