mlsanityData to 7 Oct 2026

ZipAR: Parallel Auto-regressive Image Generation through Spatial Locality

arXiv 2412.04062 ↗First posted 5 Dec 2024cs.CV cs.AIcited by 20

On the map: Efficient LLM inference and serving → Fast autoregressive visual generation

Abstract (arXiv)

In this paper, we propose ZipAR, a training-free, plug-and-play parallel decoding framework for accelerating auto-regressive (AR) visual generation. The motivation stems from the observation that images exhibit local structures, and spatially distant regions tend to have minimal interdependence. Given a partially decoded set of visual tokens, in addition to the original next-token prediction scheme in the row dimension, the tokens corresponding to spatially adjacent regions in the column dimension can be decoded in parallel, enabling the ``next-set prediction'' paradigm. By decoding multiple tokens simultaneously in a single forward pass, the number of forward passes required to generate an image is significantly reduced, resulting in a substantial improvement in generation efficiency. Experiments demonstrate that ZipAR can reduce the number of model forward passes by up to 91% on the Emu3-Gen model without requiring any additional retraining. Code is available here: https://github.com/ThisisBillhe/ZipAR.

Key facts. Cited by 20 papers in the map (data through 7 Oct 2026). Placed in Fast autoregressive visual generation, one of 12 directions in Efficient LLM inference and serving. Builds on 33 of its 39 references that are in the map. First posted on arXiv 5 Dec 2024.

Builds on

33 of 39 references.

Closest ancestors · 1

Same direction, newest first

  1. 2024Accelerating Auto-regressive Text-to-Image Generation with Training-free Speculative Jacobi Decoding2 Oct

Other references · 32

  1. 2023GPT-4 Technical Reportcited by 18,132
  2. 2014Microsoft COCO: Common Objects in Contextcited by 17,498
  3. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,082
  4. 2021Diffusion Models Beat GANs on Image Synthesiscited by 8,185
  5. 2017Neural Discrete Representation Learningcited by 4,650
  6. 2020Taming Transformers for High-Resolution Image Synthesiscited by 3,229
  7. 2013Visualizing and Understanding Convolutional Networkscited by 2,278
  8. 2022Fast Inference from Transformers via Speculative Decodingcited by 1,615
  9. 2022Scaling Autoregressive Models for Content-Rich Text-to-Image Generationcited by 1,287
  10. 2022MaskGIT: Masked Generative Image Transformercited by 1,082
  11. 2023ImageReward: Learning and Evaluating Human Preferences for Text-to-Image Generationcited by 1,010
  12. 2024Visual Autoregressive Modeling: Scalable Image Generation via Next-Scale Predictioncited by 1,005
20 more
  1. 2023Accelerating Large Language Model Decoding with Speculative Samplingcited by 975
  2. 2024Chameleon: Mixed-Modal Early-Fusion Foundation Modelscited by 904
  3. 2023Human Preference Score v2: A Solid Benchmark for Evaluating Human Preferences of Text-to-Image Synthesiscited by 847
  4. 2024Medusa: Simple LLM Inference Acceleration Framework with Multiple Decoding Headscited by 821
  5. 2024Emu3: Next-Token Prediction is All You Needcited by 761
  6. 2024Autoregressive Model Beats Diffusion: Llama for Scalable Image Generationcited by 760
  7. 2024Autoregressive Image Generation without Vector Quantizationcited by 644
  8. 2023Muse: Text-To-Image Generation via Masked Generative Transformerscited by 640
  9. 2024Evaluating Text-to-Visual Generation with Image-to-Text Generationcited by 465
  10. 2024Janus: Decoupling Visual Encoding for Unified Multimodal Understanding and Generationcited by 447
  11. 2022MAGVIT: Masked Generative Video Transformercited by 341
  12. 2024SEED-X: Multimodal Models with Unified Multi-granularity Comprehension and Generationcited by 328
  13. 2024Break the Sequential Dependency of LLM Inference Using Lookahead Decodingcited by 295
  14. 2023Accelerating LLM Inference with Staged Speculative Decodingcited by 151
  15. 2024Lumina-mGPT: Illuminate Flexible Photorealistic Text-to-Image Generation with Multimodal Generative Pretrainingcited by 148
  16. 2023Accelerating Transformer Inference for Translation via Parallel Decodingcited by 123
  17. 2023Online Speculative Decodingcited by 109
  18. 2021Nested Hierarchical Transformer: Towards Accurate, Data-Efficient and Interpretable Visual Understandingcited by 97
  19. 2024CLLMs: Consistency Large Language Modelscited by 71
  20. 2024Boosting Lossless Speculative Decoding via Feature Sampling and Partial Alignment Distillationcited by 10

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 1; 2025: 9; 2026: 10.

20 papers.

Sort:

By area:

  1. 2025MineWorld: a Real-Time and Open-Source Interactive World Model on Minecraftcited by 93
  2. 2025Neighboring Autoregressive Modeling for Efficient Visual Generationcited by 29
  3. 2024Next Patch Prediction for Autoregressive Visual Generationcited by 25
  4. 2025FastCar: Cache Attentive Replay for Fast Auto-Regressive Video Generation on the Edgecited by 9
  5. 2025Fast Autoregressive Video Generation with Diagonal Decodingcited by 8
  6. 2025XSpecMesh: Quality-Preserving Auto-Regressive Mesh Generation Acceleration via Multi-Head Speculative Decodingcited by 5
  7. 2026Multi-Scale Local Speculative Decoding for Image Generationcited by 3
  8. 2025Exploiting Discriminative Codebook Prior for Autoregressive Image Generationcited by 2
  9. 2025Generation then Reconstruction: Accelerating Masked Autoregressive Models via Two-Stage Samplingcited by 2
  10. 2025VVS: Accelerating Speculative Decoding for Visual Autoregressive Generation via Partial Verification Skippingcited by 2
  11. 2025Autoregressive Image Generation Needs Only a Few Lines of Cached Tokenscited by 2
  12. 2026Knowledge Distillation for Visual Autoregressive Modelscited by 2
8 more
  1. 2026SSD: Spatially Speculative Decoding Accelerates Autoregressive Image Generationcited by 2
  2. 2026Annealed Relaxation of Speculative Decoding for Faster Autoregressive Image Generationcited by 1
  3. 2026K-Forcing: Joint Next-K-Token Decoding via Push-Forward Language Modelingcited by 1
  4. 2026CASCADE: Context-Aware Relaxation for Speculative Image Decodingcited by 0
  5. 2026CSD: Content-aware Speculative Decoding for Efficient Image Generationcited by 0
  6. 2026Accelerating Visual On-Policy Distillation with Batched Speculative Jacobi Rolloutscited by 0
  7. 2026Is Multimodal Speculative Decoding Ready for Diffusion-Based Parallel Drafting? A Survey and Empirical Diagnosiscited by 0
  8. 2026Understanding and Exploiting Diagonal Attention Sparsity in Autoregressive Image Generationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Locality-aware Parallel Decoding for Efficient Autoregressive Image Generation
  2. 2024RandAR: Decoder-only Autoregressive Visual Generation in Random Orders
  3. 2025Autoregressive Image Generation with Randomized Parallel Decoding
  4. 2024LANTERN: Accelerating Visual Autoregressive Models with Relaxed Speculative Decoding
  5. 2025From Sequential to Spatial: Reordering Autoregression for Efficient Visual Generation
  6. 2025Hawk: Leveraging Spatial Context for Faster Autoregressive Text-to-Image Generation
  7. 2026UniGen-AR: Unifying Visual Generation with Auto-Regressive Modeling
  8. 2025SimpleAR: Pushing the Frontier of Autoregressive Visual Generation through Pretraining, SFT, and RL