๐ŸŽ“How I Study AIHISA
๐Ÿ“–Read
๐Ÿ“„Papers๐Ÿ“ฐBlogs๐ŸŽฌCourses
๐Ÿ’กLearn
๐Ÿ›ค๏ธPaths๐Ÿ“šTopics๐Ÿ’กConcepts๐ŸŽดShorts
๐ŸŽฏPractice
๐ŸงฉProblems๐ŸŽฏPrompts๐Ÿง Review
Search
How I Study AI - Learn AI Papers & Lectures the Easy Way

Papers1

AllBeginnerIntermediateAdvanced
All SourcesarXiv
#single-step denoising

DEER: Draft with Diffusion, Verify with Autoregressive Models

Intermediate
Zicong Cheng, Guo-Wei Yang et al.Dec 17arXiv

DEER is a new way to speed up big language models by letting a diffusion model draft many tokens at once and an autoregressive model double-check them.

#DEER#speculative decoding#diffusion LLM