AI Paper Atlas

Research map / Architectures and efficiency / Quantization

DLLMQuant: A Post-Training Quantization Framework Tailored for Diffusion-Based Large Language Models

Chen Xu, Zhixuan Chen, Dawei Yang

ICML 2026 · Architectures and efficiency › Quantization · approach: Post-training quantization of LLMs

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add TMAS, IA-AQ and CGQ (ablation)QuaRot · LLaDA-8B W4A4 averageaccuracy: 51.03 -> 54.29Improved
add TMAS, IA-AQ and CGQ (ablation)AWQ · LLaDA-8B W4A4 averageaccuracy: 48.09 -> 49.26Improved
add the three DLLMQuant techniquesQuaRot · GSM8K, LLaDA-8B W4A4accuracy: 44.57 -> 56.25Improved
add the three DLLMQuant techniquesQuaRot · HumanEval, LLaDA-1.5-8B W4A4accuracy: 37.33 -> 44.58Improved
add the three DLLMQuant techniquesQuaRot · Dream-7B W4A4 averageaccuracy: 58.85 -> 61.90Improved
add the three DLLMQuant techniquesAWQ · LLaDA-1.5-8B W4A4 averageaccuracy: 58.49 -> 59.94Improved
add the three DLLMQuant techniquesQuaRot · LLaDA-8B C-EVAL W4A4accuracy: 56.32 -> 61.22Improved
add IA-AQ (leave out quantizing matmul after softmax naively) (ablation)QuaRot · LLaDA-8B PIQA, cumulative MSE over stepsMSE: MSE nearly 0.1 without IA-AQImproved
add CGQ weighting by both mask state and score (ablation)QuaRot · GSM8K, 4-bitaccuracy: 44.57 -> 45.58Improved
add CGQ weighting by both mask state and score (ablation)QuaRot · HumanEval, 4-bitaccuracy: 25.33 -> 26.85Improved
use mask-only weighting in CGQ vs score & mask (ablation)QuaRot · GSM8K, 4-bitaccuracy: 45.07 -> 45.58Improved
quantize LLaDA to 4 bit weights with DLLMQuantLLaDA-8B · inference vs FP16speedup: 1.71x speedupImproved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy