Inserting Information Bottlenecks for Attribution in Transformers

12/27/2020
by   Zhiying Jiang, et al.
0

Pretrained transformers achieve the state of the art across tasks in natural language processing, motivating researchers to investigate their inner mechanisms. One common direction is to understand what features are important for prediction. In this paper, we apply information bottlenecks to analyze the attribution of each feature for prediction on a black-box model. We use BERT as the example and evaluate our approach both quantitatively and qualitatively. We show the effectiveness of our method in terms of attribution and the ability to provide insight into how information flows through layers. We demonstrate that our technique outperforms two competitive methods in degradation tests on four datasets. Code is available at https://github.com/bazingagin/IBA.

READ FULL TEXT

page 5

page 7

research
07/13/2016

A Supervised Authorship Attribution Framework for Bengali Language

Authorship Attribution is a long-standing problem in Natural Language Pr...
research
03/22/2022

A Girl Has A Name, And It's ... Adversarial Authorship Attribution for Deobfuscation

Recent advances in natural language processing have enabled powerful pri...
research
05/12/2023

Asymmetric feature interaction for interpreting model predictions

In natural language processing (NLP), deep neural networks (DNNs) could ...
research
05/03/2022

Finding patterns in Knowledge Attribution for Transformers

We analyze the Knowledge Neurons framework for the attribution of factua...
research
11/08/2022

Individualized and Global Feature Attributions for Gradient Boosted Trees in the Presence of ℓ_2 Regularization

While ℓ_2 regularization is widely used in training gradient boosted tre...
research
09/14/2022

On the State of the Art in Authorship Attribution and Authorship Verification

Despite decades of research on authorship attribution (AA) and authorshi...
research
04/09/2021

An Empirical Comparison of Instance Attribution Methods for NLP

Widespread adoption of deep models has motivated a pressing need for app...

Please sign up or login with your details

Forgot password? Click here to reset