Why Diffusion Language Models Struggle with Truly Parallel (Non-Autoregressive) Decoding?

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Li, Pengxiang, Muhtar, Dilxat, Chen, Tianlong, Yin, Lu, Liu, Shiwei
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866912929452916736
author Li, Pengxiang
Muhtar, Dilxat
Chen, Tianlong
Yin, Lu
Liu, Shiwei
author_facet Li, Pengxiang
Muhtar, Dilxat
Chen, Tianlong
Yin, Lu
Liu, Shiwei
contents Diffusion Language Models (DLMs) are often advertised as enabling parallel token generation, yet practical fast DLMs frequently converge to left-to-right, autoregressive (AR)-like decoding dynamics. In contrast, genuinely non-AR generation is promising because it removes AR's sequential bottleneck, better exploiting parallel hardware to reduce synchronization/communication overhead and improve latency scaling with output length. We argue that a primary driver of AR-like decoding is a mismatch between DLM objectives and the highly sequential structure of widely used training data, including standard pretraining corpora and long chain-of-thought (CoT) supervision. Motivated by this diagnosis, we propose NAP (Non-Autoregressive Parallel DLMs), a proof-of-concept, data-centric approach that better aligns supervision with non-AR parallel decoding. NAP curates examples as multiple independent reasoning trajectories and couples them with a parallel-forced decoding strategy that encourages multi-token parallel updates. Across math reasoning benchmarks, NAP yields stronger performance under parallel decoding than DLMs trained on standard long CoT data, with gains growing as parallelism increases. Our results suggest that revisiting data and supervision is a principled direction for mitigating AR-like behavior and moving toward genuinely non-autoregressive parallel generation in DLMs. Our code is available at https://github.com/pixeli99/NAP.
format Preprint
id arxiv_https___arxiv_org_abs_2602_23225
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Why Diffusion Language Models Struggle with Truly Parallel (Non-Autoregressive) Decoding?
Li, Pengxiang
Muhtar, Dilxat
Chen, Tianlong
Yin, Lu
Liu, Shiwei
Computation and Language
Artificial Intelligence
Diffusion Language Models (DLMs) are often advertised as enabling parallel token generation, yet practical fast DLMs frequently converge to left-to-right, autoregressive (AR)-like decoding dynamics. In contrast, genuinely non-AR generation is promising because it removes AR's sequential bottleneck, better exploiting parallel hardware to reduce synchronization/communication overhead and improve latency scaling with output length. We argue that a primary driver of AR-like decoding is a mismatch between DLM objectives and the highly sequential structure of widely used training data, including standard pretraining corpora and long chain-of-thought (CoT) supervision. Motivated by this diagnosis, we propose NAP (Non-Autoregressive Parallel DLMs), a proof-of-concept, data-centric approach that better aligns supervision with non-AR parallel decoding. NAP curates examples as multiple independent reasoning trajectories and couples them with a parallel-forced decoding strategy that encourages multi-token parallel updates. Across math reasoning benchmarks, NAP yields stronger performance under parallel decoding than DLMs trained on standard long CoT data, with gains growing as parallelism increases. Our results suggest that revisiting data and supervision is a principled direction for mitigating AR-like behavior and moving toward genuinely non-autoregressive parallel generation in DLMs. Our code is available at https://github.com/pixeli99/NAP.
title Why Diffusion Language Models Struggle with Truly Parallel (Non-Autoregressive) Decoding?
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2602.23225