GGSS：用於生成式視覺語言模型推論時去偏的測地線門控球面引導

摘要

生成式視覺語言模型（VLMs）日益廣泛應用於以人為中心的場景，然而即使影像僅在感知種族或性別等受控屬性上有所不同，它們仍可能產生具人口統計偏誤的輸出。然而，現有的推論時去偏方法大多針對靜態嵌入或類似CLIP的模型設計，而非生成式視覺語言模型。我們提出GGSS——測地線門控球面引導（Geodesic-Gated Spherical Steering）——一種保範數干預方法，在單位超球面上發現反事實偏誤子空間，沿測地線弧引導視覺標記，並利用自適應閘門專注於修正攜帶較強人口統計訊號的標記。我們在統一的單操作點協議下，針對類別、成對及職業-性別偏誤測試，評估四個生成式視覺語言模型與十種改編的推論時去偏基線及基於提示詞的緩解方法，同時衡量一般視覺語言能力。GGSS在所有四個模型上達成最低平均偏誤，其中在四個骨幹模型中的三個上，經配對置換檢驗達顯著差異，同時將MMStar準確度維持在未引導基線的±0.6個百分點以內。程式碼可在 https://github.com/dukesun99/GGSS 取得。

English

Generative vision-language models (VLMs) are increasingly used in human-centered settings, yet they can produce demographically biased outputs even when images differ only in controlled attributes such as perceived race or gender. However, existing inference-time debiasers were largely designed for static embeddings or CLIP-like models rather than generative VLMs. We propose GGSS---Geodesic-Gated Spherical Steering---a norm-preserving intervention that discovers a counterfactual bias subspace on the unit hypersphere, steers visual tokens along geodesic arcs, and uses an adaptive gate to focus correction on tokens that carry stronger demographic signal. We evaluate four generative VLMs against ten adapted inference-time debiasing baselines and prompt-based mitigation under a single operating-point protocol across categorical, pairwise, and occupation-gender bias tests, while also measuring general visual-language capability. GGSS achieves the lowest average bias on all four models, significant on three of four backbones under paired permutation tests, while preserving MMStar accuracy within +/- 0.6 p.p. of the unsteered baseline. Code is available at https://github.com/dukesun99/GGSS.