Where Does Generative Difficulty Reside? An Empirical Study of Target Representations

• URL: http://arxiv.org/abs/2608.00626v1
• Date: Sat, 01 Aug 2026 12:32:31 GMT
• Title: Where Does Generative Difficulty Reside? An Empirical Study of Target Representations
• Abstract summary: The target representation defines the distribution an image generator must learn, yet it is often treated as an interchangeable interface.<n>We study raw pixels, SD-VAE latents and DINOv2 as well as MAE representation-autoencoder features within a unified masked autoregressive rectified-flow model.<n>Our results show that compression, reconstruction fidelity, token dimensionality, and visible semantic clustering do not individually predict generative behavior.
• Score: 62.54876287800644
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The target representation defines the distribution an image generator must learn, yet it is often treated as an interchangeable interface. This assumption is particularly questionable for continuous masked generators, which combine contextual inference from visible tokens with conditional modeling of each missing token. We study raw pixels, SD-VAE latents and DINOv2 as well as MAE representation-autoencoder features within a unified masked autoregressive rectified-flow model. Under a shared ImageNet training budget, these spaces exhibit distinct optimization and inference regimes. DINOv2 converges fastest in both iterations and computation but benefits strongly from a wider local denoiser and direct context fusion. Pixels optimize substantially more slowly and require a different prediction, masking, and guidance configuration. MAE reconstructs images more faithfully and exhibits clear semantic clustering, yet produces generations substantially worse than DINOv2. The representations also respond differently to classifier-free guidance and occupy distinct precision-recall trade-offs. Together, our results show that compression, reconstruction fidelity, token dimensionality, and visible semantic clustering do not individually predict generative behavior. Instead, target representations redistribute difficulty across contextual modeling, per-token denoising, and inference-time distributional control.

Related papers

• Rarity-Aware Discrete Diffusion with Spatially Consistent Decoding for Photo-Realistic Image Super-Resolution [80.28905986756685]
  DiMOO-SR is a rarity-aware multimodal discrete diffusion framework for photo-realistic image Super-Resolution.<n>Experiments on widely used real-world SR benchmarks demonstrate that DiMOO-SR achieves competitive perceptual quality with only a few parallel decoding steps.
  arXiv  Detail & Related papers  (2026-07-20T07:01:01Z)
• Nemotron-Labs-Diffusion-Image: Advancing Masked Discrete Diffusion for High-Resolution Image Synthesis [93.7120829382642]
  Nemotron-Labs-Diffusion-Image is a state-of-the-art masked discrete diffusion model for text-to-image synthesis.<n>Standard MDMs lack self-correcting capability because discrete tokens cannot be modified once they are unmasked.<n>Nemotron-Labs-Diffusion-Image incorporates a token-editing mechanism that enables the model to dynamically revise already-unmasked tokens during inference.
  arXiv  Detail & Related papers  (2026-06-29T05:48:41Z)
• RPiAE: A Representation-Pivoted Autoencoder Enhancing Both Image Generation and Editing [37.46832944462102]
  Diffusion models have become the dominant paradigm for image generation and editing.<n>Recent attempts to leverage pretrained visual representation models as tokenizer priors either align diffusion features to representation features or directly reuse representation encoders as frozen tokenizers.<n>We propose Representation-Pivoted AutoEncoder, a representation-based tokenizer that improves both generation and editing.
  arXiv  Detail & Related papers  (2026-03-19T17:54:43Z)
• UniCom: Unified Multimodal Modeling via Compressed Continuous Semantic Representations [45.861324782424326]
  Current unified multimodal models typically rely on discrete visual tokenizers to bridge the modality gap.<n>We introduce UniCom, a unified framework that harmonizes multimodal understanding and generation via compressed continuous representation.
  arXiv  Detail & Related papers  (2026-03-11T12:14:26Z)
• ResTok: Learning Hierarchical Residuals in 1D Visual Tokenizers for Autoregressive Image Generation [64.84095852784714]
  Residual Tokenizer (ResTok) is a 1D visual tokenizer that builds hierarchical residuals for both image tokens and latent tokens.<n>We show that restoring hierarchical residual priors in visual tokenization significantly improves AR image generation, achieving a gFID of 2.34 on ImageNet-256 with only 9 sampling steps.
  arXiv  Detail & Related papers  (2026-01-07T14:09:18Z)
• Tokenize Image as a Set [17.142970970610616]
  We introduce an unordered token set representation to dynamically allocate coding capacity based on regional semantic complexity.<n>To address the challenge of modeling discrete sets, we devise a dual transformation mechanism that transforms sets into fixed-length integer sequences.<n>Experiments demonstrate our method's superiority in semantic-aware representation and generation quality.
  arXiv  Detail & Related papers  (2025-03-20T17:59:51Z)
• Frequency Autoregressive Image Generation with Continuous Tokens [31.833852108014312]
  We introduce the frequency progressive autoregressive (textbfFAR) paradigm and instantiate FAR with the continuous tokenizer.<n>We demonstrate the efficacy of FAR through comprehensive experiments on the ImageNet dataset.
  arXiv  Detail & Related papers  (2025-03-07T10:34:04Z)
• Traditional Classification Neural Networks are Good Generators: They are Competitive with DDPMs and GANs [104.72108627191041]
  We show that conventional neural network classifiers can generate high-quality images comparable to state-of-the-art generative models. We propose a mask-based reconstruction module to make semantic gradients-aware to synthesize plausible images. We show that our method is also applicable to text-to-image generation by regarding image-text foundation models.
  arXiv  Detail & Related papers  (2022-11-27T11:25:35Z)
• Diffusion Visual Counterfactual Explanations [51.077318228247925]
  Visual Counterfactual Explanations (VCEs) are an important tool to understand the decisions of an image. Current approaches for the generation of VCEs are restricted to adversarially robust models and often contain non-realistic artefacts. In this paper, we overcome this by generating Visual Diffusion Counterfactual Explanations (DVCEs) for arbitrary ImageNet classifiers.
  arXiv  Detail & Related papers  (2022-10-21T09:35:47Z)
• The Deep Generative Decoder: MAP estimation of representations improves modeling of single-cell RNA data [0.0]
  We present a simple generative model that computes model parameters and representations directly via maximum a posteriori (MAP) estimation. The advantages of this approach are its simplicity and its capability to provide representations of much smaller dimensionality than a comparable VAE.
  arXiv  Detail & Related papers  (2021-10-13T12:17:46Z)

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.