mlsanityData to 7 Oct 2026

Accelerating Diffusion LLM Inference via Local Determinism Propagation

arXiv 2510.07081 ↗First posted 8 Oct 2025cs.CLcited by 21

On the map: Efficient LLM inference and serving → Diffusion LLM and MoE inference

Abstract (arXiv)

Diffusion large language models (dLLMs) represent a significant advancement in text generation, offering parallel token decoding capabilities. However, existing open-source implementations suffer from quality-speed trade-offs that impede their practical deployment. Conservative sampling strategies typically decode only the most confident token per step to ensure quality (i.e., greedy decoding), at the cost of inference efficiency due to repeated redundant refinement iterations--a phenomenon we term delayed decoding. Through systematic analysis of dLLM decoding dynamics, we characterize this delayed decoding behavior and propose a training-free adaptive parallel decoding strategy, named LocalLeap, to address these inefficiencies. LocalLeap is built on two fundamental empirical principles: local determinism propagation centered on high-confidence anchors and progressive spatial consistency decay. By applying these principles, LocalLeap identifies anchors and performs localized relaxed parallel decoding within bounded neighborhoods, achieving substantial inference step reduction through early commitment of already-determined tokens without compromising output quality. Comprehensive evaluation on various benchmarks demonstrates that LocalLeap achieves 6.94× throughput improvements and reduces decoding steps to just 14.2% of the original requirement, achieving these gains with negligible performance impact. The source codes are available at: https://github.com/friedrichor/LocalLeap.

Key facts. Cited by 21 papers in the map (data through 7 Oct 2026). Placed in Diffusion LLM and MoE inference, one of 12 directions in Efficient LLM inference and serving. Builds on 34 of its 37 references that are in the map. First posted on arXiv 8 Oct 2025.

Builds on

34 of 37 references.

Closest ancestors · 3

Same direction, newest first

  1. 2025Sparse-dLLM: Accelerating Diffusion LLMs with Dynamic Cache Eviction4 Aug
  2. 2025Fast-dLLM: Training-free Acceleration of Diffusion LLM by Enabling KV Cache and Parallel Decoding28 May
  3. 2025dKV-Cache: The Cache for Diffusion Language Models21 May

Other references · 31

  1. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  2. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,694
  3. 2024The Llama 3 Herd of Modelscited by 14,192
  4. 2021Training Verifiers to Solve Math Word Problemscited by 9,743
  5. 2020Denoising Diffusion Implicit Modelscited by 9,655
  6. 2025Qwen3 Technical Reportcited by 8,195
  7. 2021Evaluating Large Language Models Trained on Codecited by 8,053
  8. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,005
  9. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,146
  10. 2021Measuring Mathematical Problem Solving With the MATH Datasetcited by 5,203
  11. 2021GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Modelscited by 3,595
  12. 2021Program Synthesis with Large Language Modelscited by 3,509
19 more
  1. 2023A Survey of Large Language Modelscited by 2,537
  2. 2023Stable Video Diffusion: Scaling Latent Video Diffusion Models to Large Datasetscited by 2,526
  3. 2022Video Diffusion Modelscited by 2,280
  4. 2021Structured Denoising Diffusion Models in Discrete State-Spacescited by 1,665
  5. 2023Instruction-Following Evaluation for Large Language Modelscited by 1,143
  6. 2022MaskGIT: Masked Generative Image Transformercited by 1,082
  7. 2025Large Language Diffusion Modelscited by 872
  8. 2025Dream 7B: Diffusion Large Language Modelscited by 515
  9. 2024Large Language Models: A Surveycited by 506
  10. 2024Scaling Diffusion Language Models via Adaptation from Autoregressive Modelscited by 268
  11. 2025MMaDA: Multimodal Large Diffusion Language Modelscited by 241
  12. 2025Mercury: Ultra-Fast Language Models Based on Diffusioncited by 225
  13. 2024Scaling up Masked Diffusion Models on Textcited by 195
  14. 2025Seed Diffusion: A Large-Scale Diffusion Language Model with High-Speed Inferencecited by 173
  15. 2025dLLM-Cache: Accelerating Diffusion Large Language Models with Adaptive Cachingcited by 168
  16. 2025LLaDA-V: Large Language Diffusion Models with Visual Instruction Tuningcited by 145
  17. 2025Accelerated Sampling from Masked Diffusion Models via Entropy Bounded Unmaskingcited by 128
  18. 2025LaViDa: A Large Diffusion Language Model for Multimodal Understandingcited by 81
  19. 2025Diffusion Language Models Know the Answer Before Decodingcited by 68

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 1; 2026: 20.

21 papers.

Sort:

By area:

  1. 2026DAWN: Dependency-Aware Fast Inference for Diffusion LLMscited by 12
  2. 2025Fast-Decoding Diffusion Language Models via Progress-Aware Confidence Schedulescited by 10
  3. 2026DSB: Dynamic Sliding Block Scheduling for Diffusion LLMscited by 7
  4. 2026Stop the Flip-Flop: Context-Preserving Verification for Fast Revocable Diffusion Decodingcited by 6
  5. 2026R2-dLLM: Accelerating Diffusion Large Language Models via Spatio-Temporal Redundancy Reductioncited by 6
  6. 2026DreamReasoner-8B: Block-Size Curriculum Learning for Diffusion Reasoning Modelscited by 4
  7. 2026Breaking Block Boundaries: Anchor-based History-stable Decoding for Diffusion Large Language Modelscited by 3
  8. 2026Locally Coherent Parallel Decoding in Diffusion Language Modelscited by 2
  9. 2026Cluster-Level Attention-Guided Parallel Decoding for Masked Diffusion Language Modelscited by 2
  10. 2026TACG: Trajectory-Aware Commit Gating for Diffusion Language Model Decodingcited by 2
  11. 2026GeoBlock: Inferring Block Granularity from Dependency Geometry in Diffusion Language Modelscited by 1
  12. 2026Not All Denoising Steps Are Equal: Model Scheduling for Faster Masked Diffusion Language Modelscited by 1
9 more
  1. 2026BlockBatch: Multi-Scale Consensus Decoding for Efficient Diffusion Language Model Inferencecited by 1
  2. 2026Efficient Diffusion LLMs via Temporal-Spatial Parallel Decoding and Confidence Extrapolationcited by 1
  3. 2026Supportive Token Revealing for Fast Diffusion Language Model Decodingcited by 1
  4. 2026PSD: Pushing the Pareto Frontier of Diffusion LLMs via Parallel Speculative Decodingcited by 0
  5. 2026Mean-Field Parallel Decoding for Discrete Diffusion Language Modelscited by 0
  6. 2026DC-Leap: Training-Free Acceleration of dLLMs via Draft-Guided Contiguous Leaping Decodingcited by 0
  7. 2026Affix Cache for Diffusion Large Language Modelscited by 0
  8. 2026Flash-dLLM: IO-Aware KV Caching and Parallel Decoding for Fast, Memory-Efficient Diffusion LLMscited by 0
  9. 2026Fine-Tuning Diffusion Language Models with Context Selection and Target Weightingcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Accelerating Diffusion LLMs via Adaptive Parallel Decoding
  2. 2026Early-Bird Decoding: Accelerating Diffusion LLMs with Learnable Block Sizes and Parallel Sampling
  3. 2025Free Draft-and-Verification: Toward Lossless Parallel Decoding for Diffusion Large Language Models
  4. 2026Roll Out and Roll Back: Diffusion LLMs are Their Own Efficiency Teachers
  5. 2026d3LLM: Ultra-Fast Diffusion LLM using Pseudo-Trajectory Distillation
  6. 2025From Bits to Rounds: Parallel Decoding with Exploration for Diffusion Language Models
  7. 2025Diffusion LLMs Can Do Faster-Than-AR Inference via Discrete Diffusion Forcing
  8. 2025ParallelBench: Understanding the Trade-offs of Parallel Decoding in Diffusion LLMs