AI RESEARCH
Dependency-Aware Parallel Decoding via Attention for Diffusion LLMs
arXiv CS.LG
•
ArXi:2603.12996v1 Announce Type: new Parallel decoding for diffusion LLMs (dLLMs) is difficult because each denoising step provides only token-wise marginal distributions, while unmasking multiple tokens simultaneously requires accounting for inter-token dependencies. We propose Dependency-Aware Parallel Decoding (DAPD), a simple