Researchers have proposed a new method to evaluate the quality of autoformalization tasks, which involve translating natural language statements into formal languages. The existing methods for evaluation are limited and coarse-grained, making them unsuitable for advanced formal mathematical reasoning. A systematic, automatic method is introduced that uses an ensemble of large language models (LLMs) as judges, evaluating criteria such as logical preservation, mathematical consistency, and formal quality. This approach is validated within the domain of formal mathematics and shows promise in providing a scalable, interpretable, and reliable evaluation method. --- Why it matters: This research matters to AI engineers because it addresses a crucial challenge in autoformalization: evaluating the quality of translated statements. A more accurate evaluation method can improve the reliability of AI systems in formal mathematical reasoning, which is essential for applications like automated theorem proving and proof checking. Source: https://arxiv.org/abs/2506.10903