Layer by Layer: Uncovering Hidden Representations in Language Models | Best AI papers explained | Podwise