AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

Dynamic-dLLM: Dynamic Cache-Budget and Adaptive Parallel Decoding for Training-Free Acceleration of Diffusion LLM

Tianyi Wu, Xiaoxi Sun, Yanhua Jiao, Yulin Li, Yixin Chen, Yun-Hao Cao, Yi-Qi Hu, Zhuotao Tian

ICLR 2026 · Large language models › Language, safety and interpretability · approach: Parallel and early-exit decoding · 0 citations (OpenAlex)

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
apply Dynamic-dLLM with parallel decodingLLaDA-8B-Instruct · GSM8Kthroughput (TPS): 8.32 -> 37.29Improved
apply Dynamic-dLLM with parallel decodingLLaDA-8B-Instruct · GSM8Kaccuracy: 78.21 -> 78.01No clear effect
apply Dynamic-dLLM instead of Fast-dLLM, with parallel decodingLLaDA-8B-Instruct · Average of 5 benchmarksthroughput speedup: 2.85x -> 3.21xImproved
apply Dynamic-dLLM with parallel decodingDream-v0-7B-Instruct · Average of 5 benchmarksaccuracy: 66.27 -> 65.01Hurt
increase cache budgets B_window and B_layer from 32 to 128 (ablation)Dynamic-dLLM · GSM8K (1024 tokens)accuracy: 73.92 -> 79.15Improved
remove layer budget (B_layer=0) (ablation)Dynamic-dLLM · GSM8K (1024 tokens)accuracy: 74.07 vs 77.62 at 64/64 budgetHurt
scale alpha and beta by 100x (ablation)Dynamic-dLLM · GSM8K (256 tokens)accuracy: 78.01 -> 59.76Hurt
use dynamic instead of fixed confidence threshold (ablation)Dynamic-dLLM · GSM8Kinference steps: approx. 30% fewer stepsImproved
apply Dynamic-dLLM on RTX 4090LLaDA-8B-Instruct · GSM8Kthroughput (TPS): 5.48 -> 29.77Improved

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