Implicit Modeling for Transferability Estimation of Vision Foundation Models

Yaoyan Zheng, Huiqun Wang, Nan Zhou, Di Huang

Keywords: Transferability Estimation, Model Selection

Abstract: Transferability estimation identifies the best pre-trained models for downstream tasks without incurring the high computational cost of full fine-tuning. This capability facilitates deployment and advances the pre-training and fine-tuning paradigm. However, existing methods often struggle to accurately assess transferability for emerging pre-trained models with diverse architectures, training strategies, and task alignments. In this work, we propose Implicit Transferability Modeling (ITM), a novel framework that implicitly models each model’s intrinsic transferability, coupled with a Divide-and-Conquer Variational Approximation (DVA) strategy to efficiently approximate embedding space evolution. This design enables generalization across a broader range of models and downstream tasks. Extensive experiments on a comprehensive benchmark—spanning extensive training regimes and a wider variety of model types—demonstrate that ITM consistently outperforms existing methods in terms of stability, effectiveness, and efficiency.

Primary Area: Deep learning (e.g., architectures, generative models, optimization for deep networks, foundation models, LLMs)

Submission Number: 5469