SAGE: Spuriousness-Aware Guided Prompt Exploration for Mitigating Multimodal Bias
Wenqian Ye, Di Wang, Guangtao Zheng, 刘博涵 (Bohan Liu), Aidong Zhang
摘要
此摘要由英文原文自動翻譯。
如CLIP等大型視覺-語言模型,透過在共享嵌入空間中對齊影像與文字,展現了強大的零樣本分類性能。然而,CLIP模型常發展出多模態虛假偏誤——依賴虛假特徵的不良傾向。例如,CLIP可能依據頻繁共現的背景而非物體核心特徵來推斷影像中的物體類型。此偏誤顯著損害預訓練CLIP模型在分布外資料上的強健性——在該情境下這些跨模態關聯不再成立。現有的多模態虛假偏誤緩解方法通常需要對下游資料微調或預先知道偏誤,削弱了CLIP開箱即用的可用性。本文首先對零樣本分類中多模態虛假偏誤的影響進行理論分析。基於此洞見,我們提出Spuriousness-Aware Guided Exploration(SAGE)——一個簡單有效的方法,透過引導式提示選擇緩解虛假偏誤。SAGE無需訓練、微調或外部標註。它探索提示模板空間,選擇能在類別間誘導最大語意分離的提示,從而提升最差組強健性。在四個真實世界基準資料集與五個流行主幹模型上的廣泛實驗證明,SAGE持續提升零樣本效能與泛化,在無外部知識或模型更新的情況下超越既有的零樣本方法。
原始摘要(英文)
Large vision-language models, such as CLIP, have shown strong zero-shot classification performance by aligning images and text in a shared embedding space. However, CLIP models often develop multimodal spurious biases, which is the undesirable tendency to rely on spurious features. For example, CLIP may infer object types in images based on frequently co-occurring backgrounds rather than the object's core features. This bias significantly impairs the robustness of pre-trained CLIP models on out-of-distribution data, where such cross-modal associations no longer hold. Existing methods for mitigating multimodal spurious bias typically require fine-tuning on downstream data or prior knowledge of the bias, which undermines the out-of-the-box usability of CLIP. In this paper, we first theoretically analyze the impact of multimodal spurious bias in zero-shot classification. Based on this insight, we propose Spuriousness-Aware Guided Exploration (SAGE), a simple and effective method that mitigates spurious bias through guided prompt selection. SAGE requires no training, fine-tuning, or external annotations. It explores a space of prompt templates and selects the prompts that induce the largest semantic separation between classes, thereby improving worst-group robustness. Extensive experiments on four real-world benchmark datasets and five popular backbone models demonstrate that SAGE consistently improves zero-shot performance and generalization, outperforming previous zero-shot approaches without any external knowledge or model updates.
本文貢獻
Prompt-level defense against multimodal bias: SAGE estimates how spuriousness-aware a prompt is and explores guided prompt variations to find framings where the model relies on true visual content rather than dataset shortcuts.
BibTeX
@inproceedings{ye2026sage,
title = {SAGE: Spuriousness-Aware Guided Prompt Exploration for Mitigating Multimodal Bias},
author = {Wenqian Ye and Di Wang and Guangtao Zheng and Bohan Liu and Aidong Zhang},
booktitle = {AAAI Conference on Artificial Intelligence},
year = {2026},
url = {https://arxiv.org/abs/2511.13005}
}