Building Interpretable Interaction Trees for Deep NLP Models
This paper proposes a method to disentangle and quantify interactions among words that are encoded inside a DNN for natural language processing. We construct a tree to encode salient interactions extracted by the DNN. Six metrics are proposed to analyze properties of interactions between constituents in a sentence. The interaction is defined based on Shapley values of words, which are considered as an unbiased estimation of word contributions to the network prediction. Our method is used to quantify word interactions encoded inside the BERT, ELMo, LSTM, CNN, and Transformer networks. Experimental results have provided a new perspective to understand these DNNs, and have demonstrated the effectiveness of our method.
PDF AbstractTasks
SST
CoLA
Results from the Paper
Submit
results from this paper
to get state-of-the-art GitHub badges and help the
community compare results to other papers.
Methods
Absolute Position Encodings •
Adam •
Attention Dropout •
BERT •
BiLSTM •
BPE •
Dense Connections •
Dropout •
ELMo •
GELU •
Label Smoothing •
Layer Normalization •
Linear Layer •
Linear Warmup With Linear Decay •
LSTM •
Multi-Head Attention •
Position-Wise Feed-Forward Layer •
Residual Connection •
Scaled Dot-Product Attention •
Sigmoid Activation •
Softmax •
Tanh Activation •
Transformer •
Weight Decay •
WordPiece
