NLP Compression Represents Intelligence Linearly Paper • 2404.09937 • Published Apr 15 • 27 TransformerFAM: Feedback attention is working memory Paper • 2404.09173 • Published Apr 14 • 43
