#slm-as-a-judge

Benchmark 94%의 함정, Workflow-level 평가를 통한 AI 신뢰성 확보

AI Reliability: What It Is, Why It Matters, and How to Fix It