Multi-Stage Influence Function
By Hongge Chen, Si Si, Y. Li, Ciprian Chelba, Sanjiv Kumar, Duane S. Boning, Cho-Jui Hsieh. 2020
21 citations recorded by the source corpus.
Multi-stage training and knowledge transfer, from a large-scale pretraining task to various finetuning tasks, have revolutionized natural language processing and computer vision resulting in state-of-the-art performance improvements. In this paper, we develop a multi-stage influence function score to track predictions from a finetuned model all the way back to the pretraining data. With this score, we can identify the pretraining examples in the pretraining task that contribute most to a prediction in the finetuning task. The proposed multi-stage influence function generalizes the original influence function for a single model in (Koh & Liang, 2017), thereby enabling influence computation through both pretrained and finetuned models. We study two different scenarios with the pretrained embeddings fixed or updated in the finetuning tasks. We test our proposed method in various experiments to show its effectiveness and potential applications.
Subfields: Artificial Intelligence
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Caiming Xiong
- Richard Socher
- Nazneen Rajani
- Mohit Bansal
- Yejin Choi
- Luke S. Zettlemoyer
- L. Varshney
- Yonatan Belinkov
Citations in the corpus
- BERTology Meets Biology: Interpreting Attention in Protein Language Models (2020, background): “the influential training examples for that prediction [14, 37].”
- Machine Unlearning of Features and Labels (2021, methodology): “Influence functions have then been used to trace bias in word embeddings back to documents [10, 19], determine reliable regions in learning models [46], and explain deep neural networks [7].”