Self-Attention Attribution: Interpreting Information Interactions Inside Transformer
This work addresses the need for better interpretability in Transformer-based models, particularly for understanding feature interactions, but it is incremental as it builds on existing attribution methods.
The paper tackles the problem of interpreting how input features interact within Transformer models, proposing a self-attention attribution method that identifies important attention heads for pruning with marginal performance degradation and reveals hierarchical interactions through an attribution tree.
The great success of Transformer-based models benefits from the powerful multi-head self-attention mechanism, which learns token dependencies and encodes contextual information from the input. Prior work strives to attribute model decisions to individual input features with different saliency measures, but they fail to explain how these input features interact with each other to reach predictions. In this paper, we propose a self-attention attribution method to interpret the information interactions inside Transformer. We take BERT as an example to conduct extensive studies. Firstly, we apply self-attention attribution to identify the important attention heads, while others can be pruned with marginal performance degradation. Furthermore, we extract the most salient dependencies in each layer to construct an attribution tree, which reveals the hierarchical interactions inside Transformer. Finally, we show that the attribution results can be used as adversarial patterns to implement non-targeted attacks towards BERT.