VERIFY is a large-scale dataset for enabling formal language translation. It contains formal LTL formulas with their LaTeX representations, ITL (Intermediate Technical Language) representations, and natural language translations across different domains. We used a formula enumerator to generate several LTL formulas, SPOT to verify the canonical forms of these formulas, a semantically valid rule-based approach to generate an intermediary form we call ITL, and… See the full description on the dataset page: https://huggingface.co/datasets/sedislab/VERIFY.