MUD as AI Evaluation and LLM-judge distortion in ways aggregate κ misses

(lesswrong.com) by joozio | Aug 2, 2026 | 0 comments on HN