Accelerating Diffusion LLMs via Adaptive Parallel Decoding

NeurIPSSpotlight2025

Authors: Daniel Mingyi Israel, Guy Van den Broeck, Aditya Grover
Venue: NeurIPS 2025
Track: Spotlight

TL;DR

The generation speed of LLMs are bottlenecked by autoregressive decoding, where tokens are predicted sequentially one by one…

Opening excerpt from the authors’ abstract. source

Read the paper

OpenReview

Topics

diffusion llm

← All NeurIPS 2025 Spotlight papers · Browse the whole archive