mlsanityData to 7 Oct 2026

FlexTok: Resampling Images into 1D Token Sequences of Flexible Length

arXiv 2502.13967 ↗First posted 19 Feb 2025cs.CV cs.LGcited by 105

Not on the map. Why?

Cited by 24 in Diffusion models, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

Image tokenization has enabled major advances in autoregressive image generation by providing compressed, discrete representations that are more efficient to process than raw pixels. While traditional approaches use 2D grid tokenization, recent methods like TiTok have shown that 1D tokenization can achieve high generation quality by eliminating grid redundancies. However, these methods typically use a fixed number of tokens and thus cannot adapt to an image's inherent complexity. We introduce FlexTok, a tokenizer that projects 2D images into variable-length, ordered 1D token sequences. For example, a 256x256 image can be resampled into anywhere from 1 to 256 discrete tokens, hierarchically and semantically compressing its information. By training a rectified flow model as the decoder and using nested dropout, FlexTok produces plausible reconstructions regardless of the chosen token sequence length. We evaluate our approach in an autoregressive generation setting using a simple GPT-style Transformer. On ImageNet, this approach achieves an FID<2 across 8 to 128 tokens, outperforming TiTok and matching state-of-the-art methods with far fewer tokens. We further extend the model to support to text-conditioned image generation and examine how FlexTok relates to traditional 2D tokenization. A key finding is that FlexTok enables next-token prediction to describe images in a coarse-to-fine "visual vocabulary", and that the number of tokens to generate depends on the complexity of the generation task.

Key facts. Cited by 105 papers in the map (data through 7 Oct 2026). First posted on arXiv 19 Feb 2025.

Builds on

0 of 0 references.

No references here yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 36; 2026: 69.

105 papers.

Sort:

By area:

  1. 2025GigaTok: Scaling Visual Tokenizers to 3 Billion Parameters for Autoregressive Image Generationcited by 55
  2. 2025Flow to the Mode: Mode-Seeking Diffusion Autoencoders for State-of-the-Art Image Tokenizationcited by 35
  3. 2025Selftok: Discrete Visual Tokens of Autoregression, by Diffusion, and for Reasoningcited by 31
  4. 2025"Principal Components" Enable A New Language of Imagescited by 25
  5. 2025UniFlow: A Unified Pixel Flow Tokenizer for Visual Understanding and Generationcited by 23
  6. 2025Towards Sequence Modeling Alignment between Tokenizer and Autoregressive Modelcited by 20
  7. 2025Vision Foundation Models as Effective Visual Tokenizers for Autoregressive Image Generationcited by 20
  8. 2025Lost in Latent Space: An Empirical Study of Latent Diffusion Models for Physics Emulationcited by 19
  9. 2025Latent Denoising Makes Good Tokenizerscited by 18
  10. 2025DetailFlow: 1D Coarse-to-Fine Autoregressive Image Generation via Next-Detail Predictioncited by 17
  11. 2025Learning Adaptive and Temporally Causal Video Tokenization in a 1D Latent Spacecited by 16
  12. 2026OAT: Ordered Action Tokenizationcited by 14
38 more
  1. 2025DC-AR: Efficient Masked Autoregressive Image Generation with Deep Compression Hybrid Tokenizercited by 13
  2. 2025Adapting Self-Supervised Representations as a Latent Space for Efficient Generationcited by 13
  3. 2025TokBench: Evaluating Your Visual Tokenizer before Visual Generationcited by 12
  4. 2026Mode Seeking meets Mean Seeking for Fast Long Video Generationcited by 12
  5. 2025Single-pass Adaptive Image Tokenization for Minimum Program Searchcited by 11
  6. 2026ELT: Elastic Looped Transformers for Visual Generationcited by 11
  7. 2025Robust Latent Matters: Boosting Image Generation with Sampling Error Synthesiscited by 9
  8. 2025D-AR: Diffusion via Autoregressive Modelscited by 9
  9. 2025WeTok: Powerful Discrete Tokenization for High-Fidelity Visual Reconstructioncited by 9
  10. 2026Sparse Forcing: Native Trainable Sparse Attention for Real-time Autoregressive Diffusion Video Generationcited by 9
  11. 2026Attend Before Attention: Efficient and Scalable Video Understanding via Autoregressive Gazingcited by 8
  12. 2025GPSToken: Gaussian Parameterized Spatially-adaptive Tokenization for Image Representation and Generationcited by 7
  13. 2025Decoupling High and Low Frequencies for Faithful Image Generation with Fine Detailscited by 7
  14. 2025HieraTok: Multi-Scale Visual Tokenizer Improves Image Reconstruction and Generationcited by 7
  15. 2025SSDD: Single-Step Diffusion Decoder for Efficient Image Tokenizationcited by 7
  16. 2026EVATok: Adaptive Length Video Tokenization for Efficient Visual Autoregressive Generationcited by 7
  17. 2026What Matters for Diffusion-Friendly Latent Manifold? Prior-Aligned Autoencoders for Latent Diffusioncited by 7
  18. 2025REAR: Rethinking Visual Autoregressive Models via Generator-Tokenizer Consistency Regularizationcited by 6
  19. 2025DINO-Tok: Adapting DINO for Visual Tokenizerscited by 6
  20. 2025DeRA: Decoupled Representation Alignment for Video Tokenizationcited by 6
  21. 2025InfoTok: Adaptive Discrete Video Tokenizer via Information-Theoretic Compressioncited by 6
  22. 2025VA-π: Variational Policy Alignment for Pixel-Aware Autoregressive Generationcited by 6
  23. 2026Laminating Representation Autoencoders for Efficient Diffusioncited by 6
  24. 2026SceneTok: A Compressed, Diffusable Token Space for 3D Scenescited by 6
  25. 2026Planning in 8 Tokens: A Compact Discrete Tokenizer for Latent World Modelcited by 6
  26. 2026VideoFlexTok: Flexible-Length Coarse-to-Fine Video Tokenizationcited by 6
  27. 2025ALTo: Adaptive-Length Tokenizer for Autoregressive Mask Generationcited by 5
  28. 2025Next Visual Granularity Generationcited by 5
  29. 2026Improving Flexible Image Tokenizers for Autoregressive Image Generationcited by 5
  30. 2026ResTok: Learning Hierarchical Residuals in 1D Visual Tokenizers for Autoregressive Image Generationcited by 5
  31. 2026CaTok: Taming Mean Flows for One-Dimensional Causal Image Tokenizationcited by 5
  32. 2026End-to-End Autoregressive Image Generation with 1D Semantic Tokenizercited by 5
  33. 2026LoST: Level of Semantics Tokenization for 3D Shapescited by 4
  34. 2025ProteinAE: Protein Diffusion Autoencoders for Structure Encodingcited by 3
  35. 2026Adaptive 1D Video Diffusion Autoencodercited by 3
  36. 2026One Model, Many Budgets: Elastic Latent Interfaces for Diffusion Transformerscited by 3
  37. 2026Adaptive Protein Tokenizationcited by 2
  38. 2026Probabilistic Precipitation Nowcasting with Rectified Flow Transformerscited by 2

50 most cited of 105 All 105 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2024An Image is Worth 32 Tokens for Reconstruction and Generation
  2. 2024ImageFolder: Autoregressive Image Generation with Folded Tokens
  3. 2026Mutual Enhancement Between Global Tokens and Patch Tokens: From Theory to Practice
  4. 2024Language-Guided Image Tokenization for Generation
  5. 2025Learnings from Scaling Visual Tokenizers for Reconstruction and Generation
  6. 2025CAT: Content-Adaptive Image Tokenization
  7. 2024ElasticTok: Adaptive Tokenization for Image and Video
  8. 2026Semantic-Aware Prefix Learning for Token-Efficient Image Generation