@inproceedings{lin-etal-2026-vtbench,
title = "{VTB}ench: Evaluating Visual Tokenizers for Autoregressive Image Generation",
author = "Lin, Huawei and
Geng, Tony and
Xu, Zhaozhuo and
Zhao, Weijie",
editor = "Zhang, Yue and
Xu, Ruifeng and
Purwarianti, Ayu and
Rong, Tong and
Suzuki, Jun",
booktitle = "Proceedings of the 15th International Joint Conference on Natural Language Processing and the 5th Conference of the {A}sia-Pacific Chapter of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
month = nov,
year = "2026",
address = "Hengqin, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.aacl-long.229/",
pages = "5138--5163",
ISBN = "979-8-89176-439-2",
abstract = "Autoregressive (AR) models have recently shown strong performance in image generation, where a key component is the visual tokenizer (VT) that maps continuous pixels into discrete tokens. The quality of the VT largely determines the upper bound of AR model performance. However, existing discrete VTs significantly lag behind continuous variational autoencoders (VAEs), leading to degraded reconstructions and poor preservation of fine details and text. Current benchmarks primarily evaluate end-to-end generation quality, making it difficult to isolate VT performance. We introduce VTBench, a comprehensive benchmark for systematic evaluation of visual tokenizers across three core tasks: image reconstruction, detail preservation, and text preservation. Evaluating state-of-the-art VTs with multiple metrics, we find that continuous VAEs consistently outperform discrete VTs, especially in retaining spatial structure and semantic details. Discrete VTs often produce distorted reconstructions and fail to preserve fine textures and text. We release VTBench publicly to facilitate future research on robust, general-purpose visual tokenizers."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="lin-etal-2026-vtbench">
<titleInfo>
<title>VTBench: Evaluating Visual Tokenizers for Autoregressive Image Generation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Huawei</namePart>
<namePart type="family">Lin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tony</namePart>
<namePart type="family">Geng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhaozhuo</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Weijie</namePart>
<namePart type="family">Zhao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 15th International Joint Conference on Natural Language Processing and the 5th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yue</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ruifeng</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ayu</namePart>
<namePart type="family">Purwarianti</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tong</namePart>
<namePart type="family">Rong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jun</namePart>
<namePart type="family">Suzuki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Hengqin, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-439-2</identifier>
</relatedItem>
<abstract>Autoregressive (AR) models have recently shown strong performance in image generation, where a key component is the visual tokenizer (VT) that maps continuous pixels into discrete tokens. The quality of the VT largely determines the upper bound of AR model performance. However, existing discrete VTs significantly lag behind continuous variational autoencoders (VAEs), leading to degraded reconstructions and poor preservation of fine details and text. Current benchmarks primarily evaluate end-to-end generation quality, making it difficult to isolate VT performance. We introduce VTBench, a comprehensive benchmark for systematic evaluation of visual tokenizers across three core tasks: image reconstruction, detail preservation, and text preservation. Evaluating state-of-the-art VTs with multiple metrics, we find that continuous VAEs consistently outperform discrete VTs, especially in retaining spatial structure and semantic details. Discrete VTs often produce distorted reconstructions and fail to preserve fine textures and text. We release VTBench publicly to facilitate future research on robust, general-purpose visual tokenizers.</abstract>
<identifier type="citekey">lin-etal-2026-vtbench</identifier>
<location>
<url>https://aclanthology.org/2026.aacl-long.229/</url>
</location>
<part>
<date>2026-11</date>
<extent unit="page">
<start>5138</start>
<end>5163</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T VTBench: Evaluating Visual Tokenizers for Autoregressive Image Generation
%A Lin, Huawei
%A Geng, Tony
%A Xu, Zhaozhuo
%A Zhao, Weijie
%Y Zhang, Yue
%Y Xu, Ruifeng
%Y Purwarianti, Ayu
%Y Rong, Tong
%Y Suzuki, Jun
%S Proceedings of the 15th International Joint Conference on Natural Language Processing and the 5th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2026
%8 November
%I Association for Computational Linguistics
%C Hengqin, China
%@ 979-8-89176-439-2
%F lin-etal-2026-vtbench
%X Autoregressive (AR) models have recently shown strong performance in image generation, where a key component is the visual tokenizer (VT) that maps continuous pixels into discrete tokens. The quality of the VT largely determines the upper bound of AR model performance. However, existing discrete VTs significantly lag behind continuous variational autoencoders (VAEs), leading to degraded reconstructions and poor preservation of fine details and text. Current benchmarks primarily evaluate end-to-end generation quality, making it difficult to isolate VT performance. We introduce VTBench, a comprehensive benchmark for systematic evaluation of visual tokenizers across three core tasks: image reconstruction, detail preservation, and text preservation. Evaluating state-of-the-art VTs with multiple metrics, we find that continuous VAEs consistently outperform discrete VTs, especially in retaining spatial structure and semantic details. Discrete VTs often produce distorted reconstructions and fail to preserve fine textures and text. We release VTBench publicly to facilitate future research on robust, general-purpose visual tokenizers.
%U https://aclanthology.org/2026.aacl-long.229/
%P 5138-5163
Markdown (Informal)
[VTBench: Evaluating Visual Tokenizers for Autoregressive Image Generation](https://aclanthology.org/2026.aacl-long.229/) (Lin et al., AACL-IJCNLP 2026)
ACL
- Huawei Lin, Tony Geng, Zhaozhuo Xu, and Weijie Zhao. 2026. VTBench: Evaluating Visual Tokenizers for Autoregressive Image Generation. In Proceedings of the 15th International Joint Conference on Natural Language Processing and the 5th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (Volume 1: Long Papers), pages 5138–5163, Hengqin, China. Association for Computational Linguistics.