Fetching the paper…

Nearest Neighbor Speculative Decoding for LLM Generation and Attribution · Around