DecisionPivot: a diagnostic benchmark for decision models in LLM systems with contrast pairs differing in one decisive detail, generators with independent solvers, scorer and runners
Read the original at github.com→★ 123 stars
Original headline: "StartLuxLabs/DecisionPivot: DecisionPivot: a diagnostic benchmark for decision models in LLM systems. Contrast pairs that differ in one decisive detail, generators with independent solvers, scorer and runners."