Causally Estimating the Sensitivity of Neural NLP Models to Spurious Features.
Yunxiang ZhangLiangming PanSamson TanMin-Yen KanPublished in: CoRR (2021)
Keyphrases
- classification models
- predictive power
- feature vectors
- statistical models
- neural network
- prior knowledge
- low level
- probabilistic model
- feature extraction
- natural language
- image features
- visual features
- continuous wavelet transform
- language processing
- object models
- learning rules
- machine translation
- complex systems
- false positives
- natural language processing
- machine learning