paper-with-me

Papers

Autoregressive Visual Generation Needs a Prologue

2026-05-07 · Bowen Zheng, Weijian Luo, Guang Yang, Colin Zhang, Tianyang Hu arxiv

In this work, we propose Prologue, an approach to bridging the reconstruction-generation gap in autoregressive (AR) image generation. Instead of modifying visual tokens to satisfy both reconstruction and generation, Prologue generates a small set of prologue tokens prepended to the visual token sequence. These prologue tokens are trained exclusively with the AR cross-entropy (CE) loss, while visual tokens remain dedicated to reconstruction. This decoupled design lets us optimize generation through the AR model's true distribution without affecting reconstruction quality, which we further formalize from an ELBO perspective. On ImageNet 256x256, Prologue-Base reduces gFID from 21.01 to 10.75 without classifier-free guidance while keeping reconstruction almost unchanged; Prologue-Large reaches a competitive rFID of 0.99 and gFID of 1.46 using a standard AR model without auxiliary semantic supervision. Interestingly, driven only by AR gradients, prologue tokens exhibit emergent semantic structure: linear probing on 16 prologue tokens reaches 35.88% Top-1, far above the 23.71% of the first 16 tokens from a standard tokenizer; resampling with fixed prologue tokens preserves a similar high-level semantic layout. Our results suggest a new direction: generation quality can be improved by introducing a separate learned generative representation while leaving the original representation intact.

📄 PDF Abstract BibTeX arXiv:2605.06137

Code (0)

등록된 구현이 없습니다.

Tasks

Image Generation

Similar Papers 제목 키워드 기반

Autoregressive Image Generation Needs Only a Few Lines of Cached Tokens

2025-12-04 · Ziran Qin, Youru Lv, Mingbao Lin, Zeren Zhang 외 arxiv

Autoregressive (AR) visual generation has emerged as a powerful paradigm for image and multimodal synthesis, owing to its scalability and generality. However, existing AR image generation suffers from severe memory bottl…

Text-to-Image Generation

Mirai: Autoregressive Visual Generation Needs Foresight

2026-01-21 · Yonghao Yu, Lang Huang, Zerun Wang, Runyi Li 외 arxiv

Autoregressive (AR) visual generators model images as sequences of discrete tokens and are trained with a next-token likelihood objective. This strict causal supervision optimizes each step based only on the immediate ne…

Image Generation

AR-RAG: Autoregressive Retrieval Augmentation for Image Generation

2025-06-08 · Jingyuan Qi, Zhiyang Xu, Qifan Wang, Lifu Huang

We introduce Autoregressive Retrieval Augmentation (AR-RAG), a novel paradigm that enhances image generation by autoregressively incorporating knearest neighbor retrievals at the patch level. Unlike prior methods that pe…

Image Generationparameter-efficient fine-tuningRAGRetrieval

A Survey on Vision Autoregressive Model

2024-11-13 · Kai Jiang, Jiaxing Huang

Autoregressive models have demonstrated great performance in natural language processing (NLP) with impressive scalability, adaptability and generalizability. Inspired by their notable success in NLP field, autoregressiv…

3D GenerationBenchmarkingImage GenerationMedical Image Analysis+5

UT5: Pretraining Non autoregressive T5 with unrolled denoising

2023-11-14 · Mahmoud G. Salem, Jiayu Ye, Chu-Cheng Lin, Frederick Liu

Recent advances in Transformer-based Large Language Models have made great strides in natural language generation. However, to decode K tokens, an autoregressive model needs K sequential forward passes, which may be a pe…

DenoisingQuestion GenerationQuestion-GenerationText Generation