@inproceedings{li-etal-2026-change,
title = "When to Change the Canvas? Safe Adaptive Variable-Length Decoding for Diffusion Language Models",
author = "Li, Junzhuo and
Li, Tianhao and
Wei, Hao-Ran and
Li, Xiang and
Yang, Baosong and
Hu, Xuming",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.2224/",
pages = "47641--47654",
ISBN = "979-8-89176-462-0",
abstract = "Diffusion language models generate text by denoising a masked canvas, but usually require the canvas length to be fixed before decoding. This leads to an allocation dilemma: short canvases may truncate outputs, while overlong canvases waste computation and introduce noisy continuations. Existing variable-length methods adapt the canvas through expansion or length search, but largely ignore whether such intervention is safe. We argue that canvas editing can perturb partially stabilized denoising trajectories, and propose Safe Adaptive Canvas Decoding (SACD), a training-free framework for conservative canvas control. SACD diagnoses allocation, estimates intervention risk, and chooses among expand, shrink, and keep. Experiments across reasoning, code generation, and infilling tasks show that SACD improves accuracy, latency, and token utilization across multiple DLM backbones, highlighting safe sparse intervention as a key principle for variable-length diffusion decoding."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="li-etal-2026-change">
<titleInfo>
<title>When to Change the Canvas? Safe Adaptive Variable-Length Decoding for Diffusion Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Junzhuo</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tianhao</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hao-Ran</namePart>
<namePart type="family">Wei</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiang</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Baosong</namePart>
<namePart type="family">Yang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xuming</namePart>
<namePart type="family">Hu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>Diffusion language models generate text by denoising a masked canvas, but usually require the canvas length to be fixed before decoding. This leads to an allocation dilemma: short canvases may truncate outputs, while overlong canvases waste computation and introduce noisy continuations. Existing variable-length methods adapt the canvas through expansion or length search, but largely ignore whether such intervention is safe. We argue that canvas editing can perturb partially stabilized denoising trajectories, and propose Safe Adaptive Canvas Decoding (SACD), a training-free framework for conservative canvas control. SACD diagnoses allocation, estimates intervention risk, and chooses among expand, shrink, and keep. Experiments across reasoning, code generation, and infilling tasks show that SACD improves accuracy, latency, and token utilization across multiple DLM backbones, highlighting safe sparse intervention as a key principle for variable-length diffusion decoding.</abstract>
<identifier type="citekey">li-etal-2026-change</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.2224/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>47641</start>
<end>47654</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T When to Change the Canvas? Safe Adaptive Variable-Length Decoding for Diffusion Language Models
%A Li, Junzhuo
%A Li, Tianhao
%A Wei, Hao-Ran
%A Li, Xiang
%A Yang, Baosong
%A Hu, Xuming
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F li-etal-2026-change
%X Diffusion language models generate text by denoising a masked canvas, but usually require the canvas length to be fixed before decoding. This leads to an allocation dilemma: short canvases may truncate outputs, while overlong canvases waste computation and introduce noisy continuations. Existing variable-length methods adapt the canvas through expansion or length search, but largely ignore whether such intervention is safe. We argue that canvas editing can perturb partially stabilized denoising trajectories, and propose Safe Adaptive Canvas Decoding (SACD), a training-free framework for conservative canvas control. SACD diagnoses allocation, estimates intervention risk, and chooses among expand, shrink, and keep. Experiments across reasoning, code generation, and infilling tasks show that SACD improves accuracy, latency, and token utilization across multiple DLM backbones, highlighting safe sparse intervention as a key principle for variable-length diffusion decoding.
%U https://aclanthology.org/2026.findings-emnlp.2224/
%P 47641-47654
Markdown (Informal)
[When to Change the Canvas? Safe Adaptive Variable-Length Decoding for Diffusion Language Models](https://aclanthology.org/2026.findings-emnlp.2224/) (Li et al., Findings 2026)
ACL