Admission Without Answers: Label-Free Certification and Experience Learning for LLM-Based Optimization Modeling

A method for certifying LLM-generated optimization models without ground-truth labels, potentially improving reliability of AI in mathematical modeling.

Auto-scored: This is a technical research paper on label-free certification for LLM-based optimization modeling, with no direct risk or mitigation implications under the rubric.