SELFEXPLAIN: A Self-Explaining Architecture for Neural Text Classifiers
By Dheeraj Rajagopal, Vidhisha Balachandran, Eduard H. Hovy, Yulia Tsvetkov. 2021
80 citations recorded by the source corpus.
We introduce SelfExplain, a novel self-explaining model that explains a text classifier’s predictions using phrase-based concepts. SelfExplain augments existing neural classifiers by adding (1) a globally interpretable layer that identifies the most influential concepts in the training set for a given sample and (2) a locally interpretable layer that quantifies the contribution of each local input concept by computing a relevance score relative to the predicted label. Experiments across five text-classification datasets show that SelfExplain facilitates interpretability without sacrificing performance. Most importantly, explanations from SelfExplain show sufficiency for model predictions and are perceived as adequate, trustworthy and understandable by human judges compared to existing widely-used baselines.
Subfields: Artificial Intelligence, Natural Language Processing
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Noah A. Smith
- Denny Zhou
- Yulia Tsvetkov
- Mohit Bansal
- Quoc V. Le
- Yejin Choi
- Percy Liang
- Luke S. Zettlemoyer
In Natural Language Processing
- Yulia Tsvetkov
- Noah A. Smith
- Ana Marasovic
- Mohit Bansal
- Sarah Wiegreffe
- Yejin Choi
- Kyunghyun Cho
- Graham Neubig
Citations in the corpus
- Chain of Thought Prompting Elicits Reasoning in Large Language Models (2022, background): “And while NLE aims mostly to improve neural network interpretability (Rajagopal et al., 2021), the goal of chain of thought prompting is to allow models to decompose multi-hop reasoning tasks into multiple steps—interpretability is just a side effect.”
- Measuring Association Between Labels and Free-Text Rationales (2020, background): “…has a rich history in machine learning inter-pretability as a reliable measure of rationale quality from the lens of utility to an end-user (Kim et al., 2016; Chandrasekaran et al., 2018; Hase and Bansal, 2020; Yeung et al., 2020; Poursabzi-Sangdeh et al., 2021; Rajagopal et al., 2021, i.a. ).”
- Influence Tuning: Demoting Spurious Correlations via Instance Attribution and Instance-Driven Updates (2021, background): “, 2019) and SelfExplain (Rajagopal et al., 2021) may also support instance attribution, though the flexibility of the model architecture can be more limited.”