Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Guo, Wenqi Marshall, Qian, Qingyun, Hasan, Khalad, Du, Shan
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2512.11883
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866917482485252096
author Guo, Wenqi Marshall
Qian, Qingyun
Hasan, Khalad
Du, Shan
author_facet Guo, Wenqi Marshall
Qian, Qingyun
Hasan, Khalad
Du, Shan
contents Over-aligning image generation models to a generalized aesthetic preference conflicts with user intent, particularly when "anti-aesthetic" outputs are requested for artistic or critical purposes. This adherence prioritizes developer-centered values, compromising user autonomy and aesthetic pluralism. We test this bias by constructing a wide-spectrum aesthetics dataset and evaluating state-of-the-art generation and reward models. This position paper finds that aesthetic-aligned generation models frequently default to conventionally beautiful outputs, failing to respect instructions for low-quality or negative imagery. Crucially, reward models penalize anti-aesthetic images even when they perfectly match the explicit user prompt. We confirm this systemic bias through image-to-image editing and evaluation against real abstract artworks. Our code, fine-tuned models, and datasets are available on our meta-expression intentionally anti-aesthetics webpage: https://weathon.github.io/icml2026_position/.
format Preprint
id arxiv_https___arxiv_org_abs_2512_11883
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Position: Universal Aesthetic Alignment Narrows Artistic Expression
Guo, Wenqi Marshall
Qian, Qingyun
Hasan, Khalad
Du, Shan
Computers and Society
Artificial Intelligence
Computer Vision and Pattern Recognition
Over-aligning image generation models to a generalized aesthetic preference conflicts with user intent, particularly when "anti-aesthetic" outputs are requested for artistic or critical purposes. This adherence prioritizes developer-centered values, compromising user autonomy and aesthetic pluralism. We test this bias by constructing a wide-spectrum aesthetics dataset and evaluating state-of-the-art generation and reward models. This position paper finds that aesthetic-aligned generation models frequently default to conventionally beautiful outputs, failing to respect instructions for low-quality or negative imagery. Crucially, reward models penalize anti-aesthetic images even when they perfectly match the explicit user prompt. We confirm this systemic bias through image-to-image editing and evaluation against real abstract artworks. Our code, fine-tuned models, and datasets are available on our meta-expression intentionally anti-aesthetics webpage: https://weathon.github.io/icml2026_position/.
title Position: Universal Aesthetic Alignment Narrows Artistic Expression
topic Computers and Society
Artificial Intelligence
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2512.11883