LLM Evaluation as Tensor Completion: Low-Rank Efficiency and Uncertainty Quantification - Best AI papers explained | Wave AI Podcast Notes