
research note
Two-Level Meta-Rubrics for Evaluating Open-Ended Generation: GAMUT, a Benchmark for Factual Completeness
This paper addresses the underexplored problem of evaluating factual completeness in long-form open-ended generation, a complementary dimension of.










