Towards Hierarchical Importance Attribution: Explaining Compositional Semantics for Neural Sequence Models
This work addresses the need for better interpretability in neural language models, particularly for researchers and practitioners seeking to understand and trust model decisions, though it is incremental as it builds on existing explanation methods.
The paper tackles the problem of explaining how neural sequence models handle semantic compositions by proposing hierarchical importance attribution methods, and demonstrates that their algorithms outperform prior approaches in human and metrics evaluations on LSTM and BERT models across multiple datasets.
The impressive performance of neural networks on natural language processing tasks attributes to their ability to model complicated word and phrase compositions. To explain how the model handles semantic compositions, we study hierarchical explanation of neural network predictions. We identify non-additivity and context independent importance attributions within hierarchies as two desirable properties for highlighting word and phrase compositions. We show some prior efforts on hierarchical explanations, e.g. contextual decomposition, do not satisfy the desired properties mathematically, leading to inconsistent explanation quality in different models. In this paper, we start by proposing a formal and general way to quantify the importance of each word and phrase. Following the formulation, we propose Sampling and Contextual Decomposition (SCD) algorithm and Sampling and Occlusion (SOC) algorithm. Human and metrics evaluation on both LSTM models and BERT Transformer models on multiple datasets show that our algorithms outperform prior hierarchical explanation algorithms. Our algorithms help to visualize semantic composition captured by models, extract classification rules and improve human trust of models. Project page: https://inklab.usc.edu/hiexpl/