AI Insight
This paper introduces a model recycling framework that enables transfer learning from pre-trained models without requiring access to the original training data. The framework addresses privacy concerns and data accessibility challenges by allowing practitioners to select and reuse subsets of related source models in both white-box and black-box settings. This approach enables Model as a Service providers to create libraries of efficient pre-trained models for multi-source supervised transfer learning while maintaining data privacy.
Why it matters
This framework addresses critical data privacy concerns in machine learning while maintaining model performance, making it particularly relevant for industries with strict data protection requirements like healthcare and finance. It enables more efficient and practical deployment of transfer learning by allowing organizations to leverage pre-trained models without exposing sensitive source data.
arXiv:2508.02039v2 Announce Type: replace
Abstract: Increasing concerns for data privacy and other difficulties associated with retrieving source data for model training have created the need for source-free transfer learning, in which one only has access to pre-trained models instead of data from the original source domains. This setting introduces many challenges, as many existing transfer learning methods typically rely on access to source data, which limits their direct applicability to scenarios where source data is unavailable. Further, practical concerns make it more difficult, for instance efficiently selecting models for transfer without information on source data, and transferring without full access to the source models. So motivated, we propose a model recycling framework for parameter-efficient training of models that identifies subsets of related source models to reuse in both white-box and black-box settings. Consequently, our framework makes it possible for Model as a Service (MaaS) providers to build libraries of efficient pre-trained models, thus creating an opportunity for multi-source data-free supervised transfer learning.
Source: Model Recycling Framework for Multi-Source Data-Free Supervised Transfer Learning