Using Interpretation Methods for Model Enhancement
Zhuo Chen, Chengyue Jiang, Kewei Tu
Abstract
In the age of neural natural language processing, there are plenty of works trying to derive interpretations of neural models. Intuitively, when gold rationales exist during training, one can additionally train the model to match its interpretation with the rationales. However, this intuitive idea has not been fully explored. In this paper, we propose a framework of utilizing interpretation methods and gold rationales to enhance models. Our framework is very general in the sense that it can incorporate various interpretation methods. Previously proposed gradient-based methods can be shown as an instance of our framework. We also propose two novel instances utilizing two other types of interpretation methods, erasure/replace-based and extractor-based methods, for model enhancement. We conduct comprehensive experiments on a variety of tasks. Experimental results show that our framework is effective especially in low-resource settings in enhancing models with various interpretation methods, and our two newly-proposed methods outperform gradient-based methods in most settings. Code is available at https://github. com/Chord-Chen-30/UIMER .
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
Lineage
Builds on3
- Interpretation of NLP models through input marginalizationSiwon Kim, Jihun Yi, Eunji Kim, Sungroh YoonEMNLP 2020 · 41 citations
- How do Decisions Emerge across Layers in Neural Models? Interpretation with Differentiable MaskingNicola De Cao, Michael Sejr Schlichtkrull, Wilker Aziz, Ivan TitovEMNLP 2020 · 13 citations
- Exploring Distantly-Labeled Rationales in Neural Network ModelsQuzhe Huang, Shengqi Zhu, Yansong Feng, Dongyan ZhaoACL 2021
Related papers
- A Reasoning Paradigm for Named Entity RecognitionHui Huang, Yanping Chen, Ruizhang Huang, Chuan Lin et al.AAAI 2026
- QUASER: Question Answering with Scalable Extractive RationalizationAsish Ghoshal, Srinivasan Iyer, Bhargavi Paranjape, Kushal Lakhotia et al.SIGIR 2022 · 2 citations
- Making a (Counterfactual) Difference One Rationale at a TimeMitchell Plyler, Michael Green, Min ChiNeurIPS 2021 · 12 citations
- Blocker and Matcher Can Mutually Benefit: A Co-Learning Framework for Low-Resource Entity ResolutionShiwen Wu, Qiyu Wu, Honghua Dong, Wen Hua et al.VLDB 2024 · 10 citations
- ERASER: A Benchmark to Evaluate Rationalized NLP ModelsJay DeYoung, Sarthak Jain, Nazneen Fatema Rajani, Eric P. Lehman et al.ACL 2020 · 36 citations
