# Diffusion Autoencoders are Scalable Image Tokenizers

arXiv: https://arxiv.org/abs/2501.18593 · HTML: https://mlsanity.com/paper/2501.18593/

## Key facts

- Cited by 28 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 30 Jan 2025.

## Built on by (most cited first, 28 of 28)

- [Flow to the Mode: Mode-Seeking Diffusion Autoencoders for State-of-the-Art Image Tokenization](https://mlsanity.com/paper/2503.11056/) (2025)
- ["Principal Components" Enable A New Language of Images](https://mlsanity.com/paper/2503.08685/) (2025)
- [Hyperspherical Latents Improve Continuous-Token Autoregressive Generation](https://mlsanity.com/paper/2509.24335/) (2025)
- [Image Tokenizer Needs Post-Training](https://mlsanity.com/paper/2509.12474/) (2025)
- [D-AR: Diffusion via Autoregressive Models](https://mlsanity.com/paper/2505.23660/) (2025)
- [WeTok: Powerful Discrete Tokenization for High-Fidelity Visual Reconstruction](https://mlsanity.com/paper/2508.05599/) (2025)
- [LatentUM: Unleashing the Potential of Interleaved Cross-Modal Reasoning via a Latent-Space Unified Model](https://mlsanity.com/paper/2604.02097/) (2026)
- [SSDD: Single-Step Diffusion Decoder for Efficient Image Tokenization](https://mlsanity.com/paper/2510.04961/) (2025)
- [CaTok: Taming Mean Flows for One-Dimensional Causal Image Tokenization](https://mlsanity.com/paper/2603.06449/) (2026)
- [Flow Autoencoders are Effective Protein Tokenizers](https://mlsanity.com/paper/2510.00351/) (2025)
- [VUGEN: Visual Understanding priors for GENeration](https://mlsanity.com/paper/2510.06529/) (2025)
- [ProteinAE: Protein Diffusion Autoencoders for Structure Encoding](https://mlsanity.com/paper/2510.10634/) (2025)
- [Adaptive 1D Video Diffusion Autoencoder](https://mlsanity.com/paper/2602.04220/) (2026)
- [DGAE: Diffusion-Guided Autoencoder for Efficient Latent Representation Learning](https://mlsanity.com/paper/2506.09644/) (2025)
- [Adaptive Protein Tokenization](https://mlsanity.com/paper/2602.06418/) (2026)
- [Semantic One-Dimensional Tokenizer for Image Reconstruction and Generation](https://mlsanity.com/paper/2603.16373/) (2026)
- [Probabilistic Precipitation Nowcasting with Rectified Flow Transformers](https://mlsanity.com/paper/2605.31204/) (2026)
- [Learning Linearity in Audio Consistency Autoencoders via Implicit Regularization](https://mlsanity.com/paper/2510.23530/) (2025)
- [VLIC: Vision-Language Models As Perceptual Judges for Human-Aligned Image Compression](https://mlsanity.com/paper/2512.15701/) (2025)
- [SFTok: Bridging the Performance Gap in Discrete Tokenizers](https://mlsanity.com/paper/2512.16910/) (2025)
- [ViTok-v2: Scaling Native Resolution Auto-Encoders to 5 Billion Parameters](https://mlsanity.com/paper/2605.05331/) (2026)
- [SinAE: A Single-Architecture Flow-Matching Autoencoder for Cross-Domain Atomic Systems](https://mlsanity.com/paper/2607.12380/) (2026)
- [Flow-Matching-Based Protein Structure Tokenizer Made Efficient and Easy](https://mlsanity.com/paper/2609.33129/) (2026)
- [GenDet: Painting Colored Bounding Boxes on Images via Diffusion Model for Object Detection](https://mlsanity.com/paper/2601.07273/) (2026)
- [Communication-Inspired Tokenization for Structured Image Representations](https://mlsanity.com/paper/2602.20731/) (2026)
- [RAC: Rectified Flow Auto Coder](https://mlsanity.com/paper/2603.05925/) (2026)
- [Accelerating Diffusion Decoders via Multi-Scale Sampling and One-Step Distillation](https://mlsanity.com/paper/2603.19570/) (2026)
- [LoopVAE: Recurrent Depth Across Scales for Visual Tokenization](https://mlsanity.com/paper/2609.11516/) (2026)

## Abstract (arXiv)

Tokenizing images into compact visual representations is a key step in learning efficient and high-quality image generative models. We present a simple diffusion tokenizer (DiTo) that learns compact visual representations for image generation models. Our key insight is that a single learning objective, diffusion L2 loss, can be used for training scalable image tokenizers. Since diffusion is already widely used for image generation, our insight greatly simplifies training such tokenizers. In contrast, current state-of-the-art tokenizers rely on an empirically found combination of heuristics and losses, thus requiring a complex training recipe that relies on non-trivially balancing different losses and pretrained supervised models. We show design decisions, along with theoretical grounding, that enable us to scale DiTo for learning competitive image representations. Our results show that DiTo is a simpler, scalable, and self-supervised alternative to the current state-of-the-art image tokenizer which is supervised. DiTo achieves competitive or better quality than state-of-the-art in image reconstruction and downstream image generation tasks.