Skeleton-of-Thought: Large Language Models Can Do Parallel Decoding Paper • 2307.15337 • Published Jul 28, 2023 • 36