arXiv cs.CY· Changheon Han, Hugh Xuechen Liu, Ashkan Panahi, K{\i}van\c{c} Tatar·· 11 小时前AI 评分31
超越相似性：音乐生成模型训练数据归因的层级分解
Functional Dependency beyond Similarity: Hierarchical Decomposition of Training Data Attribution in Music Generative Models
AI 导读
研究提出将训练数据归因（TDA）分数分解为全局、组和特定三个层级，以解决少数训练实例对几乎所有生成结果都排名最高的全局影响问题。在多个音乐生成模型和TDA方法上，全局与组级影响均不可忽略。作者据此提出"功能依赖"概念，区分不同范围的数据依赖关系，使TDA分数可在法律、经济和伦理的报酬讨论中解读为不同形式的贡献。
正文 · 原文
Abstract:Copyright analysis focuses on the similarity between two works. However, generative AI models learn from training data through functional relations beyond similarity. Training Data Attribution (TDA) estimates the influence of each training instance on a generated output and may serve as a basis for remuneration. However, prior studies find global influence, where a few training instances rank as the most influential for almost every generation. Used directly, TDA scores would distribute part of every remuneration to these instances regardless of what was generated. To address this, we propose a new approach in which TDA scores are decomposed into three levels of influence, namely global indicating the influence of a training data to a trained model, group for its influence on a group of generated outputs (that could be clustered together), and specific for its influence on a single generated output. Our study finds that, across several music generative models and TDA methods, global- and group-level influence is non-negligible in every case. Based on these observations, we propose the term functional dependency for the relation of training data to a trained model or to its outputs. This notion both encompasses and distinguishes dependencies of different scopes, allowing TDA scores to be interpreted as distinct forms of contribution in legal, economic and ethical discussions on remuneration.
| Comments: | 23 pages, 9 figures, 11 tables. Under review |
| Subjects: | Computers and Society (cs.CY); Machine Learning (cs.LG) |
| ACM classes: | K.5.1; I.2.6; H.5.5 |
| Cite as: | arXiv:2611.08839 [cs.CY] |
| (or arXiv:2611.08839v1 [cs.CY] for this version) | |
| https://doi.org/10.48550/arXiv.2611.08839 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: ChangHeon Han [view email]
[v1]
Fri, 6 Nov 2026 00:53:52 UTC (395 KB)
来源：arXiv cs.CY · arxiv.org