Adaptive Decoding Hits 4.2x, Joint Training 10x Faster

13 selected from 376 papers

Featured

Also Worth Noting