Converting formal logic into natural language automatically

FOL2NS addresses the gap between formal logic and natural language by generating synthetic training data for semantic parsing and theorem validation. The framework combines rule-driven modules with fine-tuned language models to produce diverse, well-formed logical expressions and their natural language counterparts. Experiments show the system generates fluent statements reliably on simpler structures, but struggles with semantic precision and natural phrasing as logical complexity—particularly nested quantifiers—increases. This work targets researchers building semantic parsing systems and formal reasoning pipelines.