SurrogateSHAP:文本到图像(T2I)模型的无训练贡献者归因
SurrogateSHAP: Training-Free Contributor Attribution for Text-to-Image (T2I) Models
AI总结:
针对文本到图像扩散模型中数据贡献者公平估值的高计算成本问题,提出基于预训练模型推理的无重训练框架SurrogateSHAP,利用梯度提升树近似效用函数并解析计算Shapley值,在多个任务上以更低开销超越现有方法。
摘要:
随着文本到图像(T2I)扩散模型在现实创意工作流中的广泛应用,一个用于评估提供数据集合的贡献者的原则性框架对于公平补偿和可持续数据市场至关重要。虽然Shapley值提供了理论上有依据的归因方法,但它面临双重计算瓶颈:(i)对每个采样的玩家(即数据贡献者)子集进行穷举模型重训练的高昂成本,以及(ii)由于贡献者交互,估计边际贡献所需的子集组合数量巨大。为此,我们提出了SurrogateSHAP,一个无需重训练的框架,通过从预训练模型进行推理来近似昂贵的重训练博弈。为了进一步提高效率,我们采用梯度提升树来近似效用函数,并基于树模型解析地推导Shapley值。我们在三个不同的归因任务上评估了SurrogateSHAP:(i)CIFAR-20上DDPM-CFG的图像质量,(ii)后印象派艺术品上Stable Diffusion的美学质量,以及(iii)时尚产品数据上FLUX.1的产品多样性。在各种设置下,SurrogateSHAP在显著降低计算开销的同时优于先前方法,一致地在多个效用指标上识别出有影响力的贡献者。最后,我们证明了SurrogateSHAP能够有效定位导致临床图像中虚假相关的数据源,为审计安全关键型生成模型提供了一条可扩展的路径。
英文摘要:
As Text-to-Image (T2I) diffusion models are increasingly used in real-world creative workflows, a principled framework for valuing contributors who provide a collection of data is essential for fair compensation and sustainable data marketplaces. While the Shapley value offers a theoretically grounded approach to attribution, it faces a dual computational bottleneck: (i) the prohibitive cost of exhaustive model retraining for each sampled subset of players (i.e., data contributors) and (ii) the combinatorial number of subsets needed to estimate marginal contributions due to contributor interactions. To this end, we propose SurrogateSHAP, a retraining-free framework that approximates the expensive retraining game through inference from a pretrained model. To further improve efficiency, we employ a gradient-boosted tree to approximate the utility function and derive Shapley values analytically from the tree-based model. We evaluate SurrogateSHAP across three diverse attribution tasks: (i) image quality for DDPM-CFG on CIFAR-20, (ii) aesthetics for Stable Diffusion on Post-Impressionist artworks, and (iii) product diversity for FLUX.1 on Fashion-Product data. Across settings, SurrogateSHAP outperforms prior methods while substantially reducing computational overhead, consistently identifying influential contributors across multiple utility metrics. Finally, we demonstrate that SurrogateSHAP effectively localizes data sources responsible for spurious correlations in clinical images, providing a scalable path toward auditing safety-critical generative models.
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。