位置：通用审美对齐限制了艺术表达
Position: Universal Aesthetic Alignment Narrows Artistic Expression
AI总结：
本文探讨了将图像生成模型过度对齐通用审美偏好与用户意图冲突的问题，指出生成模型倾向于产生传统美观输出，忽视用户对低质量或负面图像的指令。
摘要：
过度将图像生成模型对齐于通用审美偏好与用户意图相冲突，尤其是在请求“反审美”输出用于艺术或批评目的时。这种倾向优先考虑开发者价值观，损害了用户自主权和审美多样性。我们通过构建宽谱审美数据集并评估最先进的生成和奖励模型来测试这一偏差。本文发现，审美对齐的生成模型经常默认产生传统上美观的输出，未能尊重用户对低质量或负面图像的指令。关键的是，奖励模型即使在完美匹配用户显式提示时也会惩罚反审美图像。我们通过图像到图像编辑和与真实抽象艺术品的评估确认了这种系统偏差。我们的代码、微调模型和数据集可在我们的meta-expression故意反审美网页上获得：https://weathon.github.io/icml2026_position/
英文摘要：
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/.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
