Refining Language Models with Compositional Explanations

By Huihan Yao, Ying Chen, Qinyuan Ye, Xisen Jin, Xiang Ren. 2021

38 citations recorded by the source corpus.

Pre-trained language models have been successful on text classification tasks, but are prone to learning spurious correlations from biased datasets, and are thus vulnerable when making inferences in a new domain. Prior work reveals such spurious patterns via post-hoc explanation algorithms which compute the importance of input features. Further, the model is regularized to align the importance scores with human knowledge, so that the unintended model behaviors are eliminated. However, such a regularization technique lacks flexibility and coverage, since only importance scores towards a pre-defined list of features are adjusted, while more complex human knowledge such as feature interaction and pattern generalization can hardly be incorporated. In this work, we propose to refine a learned language model for a target domain by collecting human-provided compositional explanations regarding observed biases. By parsing these explanations into executable logic rules, the human-specified refinement advice from a small set of explanations can be generalized to more training examples. We additionally introduce a regularization term allowing adjustments for both importance and interaction of features to better rectify model behavior. We demonstrate the effectiveness of the proposed approach on two text classification tasks by showing improved performance in target domain as well as improved model fairness after refinement.

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

1. Denny Zhou
2. Quoc V. Le
3. Xuezhi Wang
4. Jason Wei
5. Luke S. Zettlemoyer
6. Ed H. Chi
7. Greg Durrett
8. Mohit Bansal

In Natural Language Processing

1. Greg Durrett
2. Denny Zhou
3. Jason Wei
4. Xi Ye
5. Dale Schuurmans
6. Ed H. Chi
7. Francesca Toni
8. Xuezhi Wang

Citations in the corpus

• Chain of Thought Prompting Elicits Reasoning in Large Language Models (2022, background): “As examples, it has been shown that natural language intermediate steps can improve performance (Zaidan et al., 2007; Yao et al., 2021; Hase and Bansal, 2022; Gu et al., 2022), improve robustness (Chen et al.”
• The Unreliability of Explanations in Few-shot Prompting for Textual Reasoning (2022, background): “Another line of work seeks to align the reasoning process of a trained model with the explanations, which is typically done by interpreting a prediction post-hoc through explanation techniques and optimizing the distance between the obtained explanation and ground truth explanation (Liu and Avci, 20”
• Explanation-Based Human Debugging of NLP Models: A Survey (2021, background): “C L ] 2 5 Se p 20 21 Graliński et al., 2019), while others stress that debugging must not only reveal the causes of problems but also fix or mitigate them (Kulesza et al., 2015; Yousefzadeh and O’Leary, 2019).”
• The Unreliability of Explanations in Few-Shot In-Context Learning (2022, background): “Another line of work seeks to aligning the reasoning process of a trained models with the explanations, which is typically done by interpreting a prediction post-hoc through explanation techniques and optimize the distance between the obtained explanation and ground truth explanation (Liu and Avci, ”