Wang, J., Lin, X., Qiao, R., Foo, C. & Low, B.K.H.. (2024). Helpful or Harmful Data? Fine-tuning-free Shapley Attribution for Explaining Language Model Predictions. Proceedings of the 41st International Conference on Machine Learning, in Proceedings of Machine Learning Research 235:50960-50991
Abstract:
The increasing complexity of foundational models underscores the necessity for explainability, particularly for fine-tuning, the most widely used training method for adapting models to downstream tasks. Instance attribution, one type of explanation, attributes the model prediction to each training example by an instance score. However, the robustness of instance scores, specifically towards dataset resampling, has been overlooked. To bridge this gap, we propose a notion of robustness on the sign of the instance score. We theoretically and empirically demonstrate that the popular leave-one-out-based methods lack robustness, while the Shapley value behaves significantly better, but at a higher computational cost. Accordingly, we introduce an efficient fine-tuning-free approximation of the Shapley value (FreeShap) for instance attribution based on the neural tangent kernel. We empirically demonstrate that FreeShap outperforms other methods for instance attribution and other data-centric applications such as data removal, data selection, and wrong label detection, and further generalize our scale to large language models (LLMs). Our code is available at https://github.com/JTWang2000/FreeShap.
License type:
Attribution-NonCommercial 4.0 International (CC BY-NC 4.0)
Funding Info:
This research / project is supported by the National Research Foundation - AI Singapore Programme
Grant Reference no. : AISG2-PhD/2021-08-017[T]
This research / project is supported by the National Research Foundation and Singapore Ministry of Digital Development and Innovation, National AI Group - AI Visiting Professorship Programme
Grant Reference no. : AIVP-2024-001