Variable-Length Tokenization via Learnable Global Merging for Diffusion Transformers

• URL: http://arxiv.org/abs/2606.20076v1
• Date: Thu, 18 Jun 2026 10:52:08 GMT
• Title: Variable-Length Tokenization via Learnable Global Merging for Diffusion Transformers
• Abstract summary: We propose a variable-length tokenizer that modulates length by merging tokens.<n>We show that encouraging similar tokens to merge enables direct cross-length representation alignment.<n>On ImageNet 256$times $256 generation, our merging-based variable-length tokenizer integrated with a diffusion transformer achieves a superior gFID- compute trade-off.
• Score: 20.363335535997432
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Latent Diffusion Models (LDMs) have become dominant in visual synthesis, but their quality-compute trade-off is largely constrained by the tokenizer's fixed compression ratio. Variable-length tokenizers (VLTs) promise adaptive compression by varying token counts, allowing diffusion models to flexibly balance quality and compute. However, conventional VLTs modulate length by truncating ordered token sequences, which makes token semantics depend on token position and breaks representational alignment across lengths. This leads to a cross-length shift in the latent distribution that hinders a single variable-length diffusion model from operating effectively. To address this, we propose a novel variable-length tokenizer that modulates length by merging tokens. We show that encouraging similar tokens to merge enables direct cross-length representation alignment when the diffusion transformer operates according to the merging pattern. Since conventional merging methods are data-dependent, making the merging pattern inaccessible during generation, we introduce learnable global merging, which is data-independent, to ensure compatibility with diffusion transformers. On ImageNet 256$\times$256 generation, our merging-based variable-length tokenizer integrated with a diffusion transformer achieves a superior gFID-compute trade-off compared to prior VLT methods. Code is available at [this https URL](https://github.com/movinghoon/lgm)

Related papers

• Representation-based Masked Diffusion Model [3.3928176963185237]
  Masked Diffusion Models (MDMs) have emerged as a compelling paradigm for language modeling, offering the capability for efficient parallel text generation.<n>We propose Representation-based Masked Diffusion Model (RMDM), a framework that leverages the text representation to explicitly encode global semantics.
  arXiv  Detail & Related papers  (2026-09-11T03:00:52Z)
• One Model, Many Budgets: Elastic Latent Interfaces for Diffusion Transformers [80.19461768457622]
  Elastic Latent Interface Transformer (ELIT) is a drop-in, DiT-compatible mechanism that decouples input image size from compute.<n>Read and Write cross-attention layers move information between spatial tokens and latents and prioritize important input regions.<n>On ImageNet-1K 512px, ELIT delivers an average gain of $35.3%$ and $39.6%$ in FID and FDD scores.
  arXiv  Detail & Related papers  (2026-03-12T17:57:04Z)
• Adaptation to Intrinsic Dependence in Diffusion Language Models [5.185131234265025]
  Diffusion language models (DLMs) have emerged as a promising alternative to autoregressive (AR) approaches.<n>We introduce a distribution-agnostic unmasking schedule for DLMs that adapts to the (unknown) dependence structure of the target data distribution.<n>Our results significantly improve upon prior convergence theories and yield substantial sampling acceleration for low-complexity distributions.
  arXiv  Detail & Related papers  (2026-02-23T18:41:34Z)
• Less Is More: Generating Time Series with LLaMA-Style Autoregression in Simple Factorized Latent Spaces [29.349850914912846]
  We propose FAR-TS, a simple yet effective framework that combines disentangled factorization with an autoregressive Transformer over a discrete, quantized latent space to generate time series.<n>A LLaMA-style autoregressive Transformer then models these token sequences, enabling fast and controllable generation of sequences with arbitrary length.
  arXiv  Detail & Related papers  (2025-11-07T04:15:38Z)
• A Transformer Inspired AI-based MIMO receiver [0.5039813366558306]
  The AttDet design combines model-based interpretability with data-driven flexibility.<n>We demonstrate through link-level simulations under 5G channel models and high-order, mixed QAM modulation and coding schemes.<n>AttDet can approach near-optimal BER/BLER performance while maintaining predictable, realistic complexity.
  arXiv  Detail & Related papers  (2025-10-23T09:05:10Z)
• Accelerating Diffusion LLMs via Adaptive Parallel Decoding [60.407727995313074]
  We introduce adaptive parallel decoding (APD), a novel method that dynamically adjusts the number of tokens sampled in parallel.<n>APD provides markedly higher throughput with minimal quality degradations on downstream benchmarks.
  arXiv  Detail & Related papers  (2025-05-31T06:10:10Z)
• Multimodal Latent Language Modeling with Next-Token Diffusion [111.93906046452125]
  Multimodal generative models require a unified approach to handle both discrete data (e.g., text and code) and continuous data (e.g., image, audio, video)<n>We propose Latent Language Modeling (LatentLM), which seamlessly integrates continuous and discrete data using causal Transformers.
  arXiv  Detail & Related papers  (2024-12-11T18:57:32Z)
• ACDiT: Interpolating Autoregressive Conditional Modeling and Diffusion Transformer [95.80384464922147]
  ACDiT is a blockwise Conditional Diffusion Transformer.<n>It offers a flexible between token-wise autoregression and full-sequence diffusion.<n>We show that ACDiT performs best among all autoregressive baselines on image and video generation tasks.
  arXiv  Detail & Related papers  (2024-12-10T18:13:20Z)
• Token Fusion: Bridging the Gap between Token Pruning and Token Merging [71.84591084401458]
  Vision Transformers (ViTs) have emerged as powerful backbones in computer vision, outperforming many traditional CNNs. computational overhead, largely attributed to the self-attention mechanism, makes deployment on resource-constrained edge devices challenging. We introduce "Token Fusion" (ToFu), a method that amalgamates the benefits of both token pruning and token merging.
  arXiv  Detail & Related papers  (2023-12-02T04:29:19Z)
• Adaptive Fourier Neural Operators: Efficient Token Mixers for Transformers [55.90468016961356]
  We propose an efficient token mixer that learns to mix in the Fourier domain. AFNO is based on a principled foundation of operator learning. It can handle a sequence size of 65k and outperforms other efficient self-attention mechanisms.
  arXiv  Detail & Related papers  (2021-11-24T05:44:31Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.