What's this about?
OR-Clarify tests whether a model notices missing objectives, constraints or business rules before turning a natural-language request into an optimisation problem. The proposed InterOPT system decides whether to ask another question or stop and formulate the model.
What exactly is a “silent assumption”?
It is a missing requirement the agent fills in without telling the user — such as minimising cost when the real priority was delivery time. The resulting mathematics can be correct for the wrong problem.
Why is this worth noting?
Clarification is a core agent skill that clean benchmarks usually erase. Measuring missing-slot recovery and interaction cost rewards asking the smallest useful question rather than blindly proceeding.