mlsanityData to 14 Oct 2026

Soft Tail-dropping for Adaptive Visual Tokenization

arXiv 2601.14246 ↗First posted 20 Jan 2026cs.CVcited by 1

Not on the map. Why?

Abstract (arXiv)

We present Soft Tail-dropping Adaptive Tokenizer (STAT), a 1D discrete visual tokenizer that adaptively chooses the number of output tokens per image according to its structural complexity and level of detail. STAT encodes an image into a sequence of discrete codes together with per-token keep probabilities. Beyond standard autoencoder objectives, we regularize these keep probabilities to be monotonically decreasing along the sequence and explicitly align their distribution with an image-level complexity measure. As a result, STAT produces length-adaptive 1D visual tokens that are naturally compatible with causal 1D autoregressive (AR) visual generative models. On ImageNet-1k, equipping vanilla causal AR models with STAT yields competitive or superior visual generation quality compared to other probabilistic model families, while also exhibiting favorable scaling behavior that has been elusive in prior vanilla AR visual generation attempts.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 62 of its 71 references that are in the map. First posted on arXiv 20 Jan 2026.

Builds on

62 of 71 references.

References · 62

  1. 2017Attention Is All You Needcited by 60,361
  2. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  3. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  4. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  5. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 14,132
  6. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,177
  7. 2020Denoising Diffusion Implicit Modelscited by 9,781
  8. 2021Diffusion Models Beat GANs on Image Synthesiscited by 8,255
  9. 2022Scalable Diffusion Models with Transformerscited by 6,054
  10. 2022Classifier-Free Diffusion Guidancecited by 5,847
  11. 2017Neural Discrete Representation Learningcited by 4,712
  12. 2021RoFormer: Enhanced Transformer with Rotary Position Embeddingcited by 4,703
38 more
  1. 2021Zero-Shot Text-to-Image Generationcited by 4,288
  2. 2022LAION-5B: An open large-scale dataset for training next generation image-text modelscited by 4,152
  3. 2016Gaussian Error Linear Units (GELUs)cited by 3,502
  4. 2020Taming Transformers for High-Resolution Image Synthesiscited by 3,253
  5. 2016Improved Techniques for Training GANscited by 2,970
  6. 2015Neural Machine Translation of Rare Words with Subword Unitscited by 2,918
  7. 2018Large Scale GAN Training for High Fidelity Natural Image Synthesiscited by 2,626
  8. 2019Generating Diverse High-Fidelity Images with VQ-VAE-2cited by 1,387
  9. 2018Towards Accurate Generative Models of Video: A New Metric & Challengescited by 1,234
  10. 2022MaskGIT: Masked Generative Image Transformercited by 1,099
  11. 2024Visual Autoregressive Modeling: Scalable Image Generation via Next-Scale Predictioncited by 1,023
  12. 2023GenEval: An Object-Focused Framework for Evaluating Text-to-Image Alignmentcited by 972
  13. 2019Improved Precision and Recall Metric for Assessing Generative Modelscited by 944
  14. 2025Emerging Properties in Unified Multimodal Pretrainingcited by 897
  15. 2025Cosmos World Foundation Model Platform for Physical AIcited by 864
  16. 2024Autoregressive Model Beats Diffusion: Llama for Scalable Image Generationcited by 777
  17. 2024SiT: Exploring Flow and Diffusion-based Generative Models with Scalable Interpolant Transformerscited by 737
  18. 2022Autoregressive Image Generation using Residual Quantizationcited by 722
  19. 2021Vector-quantized Image Modeling with Improved VQGANcited by 658
  20. 2024Autoregressive Image Generation without Vector Quantizationcited by 651
  21. 2023Language Model Beats Diffusion -- Tokenizer is Key to Visual Generationcited by 640
  22. 2022StyleGAN-XL: Scaling StyleGAN to Large Diverse Datasetscited by 492
  23. 2023Scaling up GANs for Text-to-Image Synthesiscited by 492
  24. 2023Finite Scalar Quantization: VQ-VAE Made Simplecited by 440
  25. 2024An Image is Worth 32 Tokens for Reconstruction and Generationcited by 302
  26. 2023MDTv2: Masked Diffusion Transformer is a Strong Image Synthesizercited by 262
  27. 2022Long Video Generation with Time-Agnostic VQGAN and Time-Sensitive Transformercited by 244
  28. 2024Open-MAGVIT2: An Open-Source Project Toward Democratizing Auto-regressive Visual Generationcited by 155
  29. 2024Randomized Autoregressive Visual Generationcited by 127
  30. 2024OmniTokenizer: A Joint Image-Video Tokenizer for Visual Generationcited by 118
  31. 2024Image and Video Tokenization with Binary Spherical Quantizationcited by 112
  32. 2025FlexTok: Resampling Images into 1D Token Sequences of Flexible Lengthcited by 110
  33. 2014Learning Ordered Representations with Nested Dropoutcited by 84
  34. 2024RandAR: Decoder-only Autoregressive Visual Generation in Random Orderscited by 83
  35. 2025Beyond Next-Token: Next-X Prediction for Autoregressive Visual Generationcited by 71
  36. 2022Improved Masked Image Generation with Token-Criticcited by 64
  37. 2025Democratizing Text-to-Image Masked Generative Models with Compact Text-Aware One-Dimensional Tokenscited by 50
  38. 2025One-D-Piece: Image Tokenizer Meets Quality-Controllable Compressioncited by 48

50 most cited of 62 All 62 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026ResAdapt: Adaptive Resolution for Efficient Multimodal Reasoningcited by 8