paper-with-me

홈 › Papers

Highly Compressed Tokenizer Can Generate Without Training

2025-06-09 · L. Lao Beyer, T. Li, X. Chen, S. Karaman, K. He

Commonly used image tokenizers produce a 2D grid of spatially arranged tokens. In contrast, so-called 1D image tokenizers represent images as highly compressed one-dimensional sequences of as few as 32 discrete tokens. We find that the high degree of compression achieved by a 1D tokenizer with vector quantization enables image editing and generative capabilities through heuristic manipulation of tokens, demonstrating that even very crude manipulations -- such as copying and replacing tokens between latent representations of images -- enable fine-grained image editing by transferring appearance and semantic attributes. Motivated by the expressivity of the 1D tokenizer's latent space, we construct an image generation pipeline leveraging gradient-based test-time optimization of tokens with plug-and-play loss functions such as reconstruction or CLIP similarity. Our approach is demonstrated for inpainting and text-guided image editing use cases, and can generate diverse and realistic samples without requiring training of any generative model.

📄 PDF Abstract BibTeX arXiv:2506.08257

Code (1)

lukaslaobeyer/token-opt 공식 구현 jax

Tasks

Image GenerationQuantizationtext-guided-image-editing

Methods 이 논문이 사용한 방법론

CLIP Contrastive Language-Image Pre-training (CLIP), consisting of a simplified version of ConVIRT trained from scratch, is an efficient method of image representation learning…
Inpainting Train a convolutional neural network to generate the contents of an arbitrary image region conditioned on its surroundings.

Similar Papers 제목 키워드 기반

Training LLMs over Neurally Compressed Text

2024-04-04 · Brian Lester, Jaehoon Lee, Alex Alemi, Jeffrey Pennington 외

In this paper, we explore the idea of training large language models (LLMs) over highly compressed text. While standard subword tokenizers compress text by a small factor, neural text compressors can achieve much higher …

Progressive Growing of Video Tokenizers for Highly Compressed Latent Spaces

2025-01-09 · Aniruddha Mahapatra, Long Mai, Yitian Zhang, David Bourgin 외

Video tokenizers are essential for latent video diffusion models, converting raw video data into spatiotemporally compressed latent spaces for efficient training. However, extending state-of-the-art video tokenizers to a…

Video Generation

Proxy Compression for Language Modeling

2026-02-04 · Lin Zheng, Xinyu Li, Qian Liu, Xiachong Feng 외 arxiv

Modern language models are trained almost exclusively on token sequences produced by a fixed tokenizer, an external lossless compressor often over UTF-8 byte sequences, thereby coupling the model to that compressor. This…

Training Language Models with homotokens Leads to Delayed Overfitting

2026-01-06 · Adrian Cosma, Stefan Ruseti, Emilian Radoi, Mihai Dascalu arxiv

Subword tokenization introduces a computational layer in language models where many distinct token sequences decode to the same surface form and preserve meaning, yet induce different internal computations. Despite this …

Data Augmentation

SceneTok: A Compressed, Diffusable Token Space for 3D Scenes

2026-02-21 · Mohammad Asim, Christopher Wewer, Jan Eric Lenssen arxiv

We present SceneTok, a novel tokenizer for encoding view sets of scenes into a compressed and diffusable set of unstructured tokens. Existing approaches for 3D scene representation and generation commonly use 3D data str…

Scene Generation