# Teaching Machine Comprehension with Compositional Explanations

arXiv: https://arxiv.org/abs/2005.00806 · HTML: https://mlsanity.com/paper/2005.00806/

## Key facts

- Cited by 22 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 2 May 2020.

## Built on by (most cited first, 22 of 22)

- [Explaining Answers with Entailment Trees](https://mlsanity.com/paper/2104.08661/) (2021)
- [Teach Me to Explain: A Review of Datasets for Explainable Natural Language Processing](https://mlsanity.com/paper/2102.12060/) (2021)
- [State-of-the-art generalisation research in NLP: A taxonomy and review](https://mlsanity.com/paper/2210.03050/) (2022)
- [ExplaGraphs: An Explanation Graph Generation Task for Structured Commonsense Reasoning](https://mlsanity.com/paper/2104.07644/) (2021)
- [e-CARE: a New Dataset for Exploring Explainable Causal Reasoning](https://mlsanity.com/paper/2205.05849/) (2022)
- [Large Language Model Instruction Following: A Survey of Progresses and Challenges](https://mlsanity.com/paper/2303.10475/) (2023)
- [On the Diversity and Limits of Human Explanations](https://mlsanity.com/paper/2106.11988/) (2021)
- [Refining Language Models with Compositional Explanations](https://mlsanity.com/paper/2103.10415/) (2021)
- [METGEN: A Module-Based Entailment Tree Generation Framework for Answer Explanation](https://mlsanity.com/paper/2205.02593/) (2022)
- ["Why is this misleading?": Detecting News Headline Hallucinations with Explanations](https://mlsanity.com/paper/2302.05852/) (2023)
- [multiPRover: Generating Multiple Proofs for Improved Interpretability in Rule Reasoning](https://mlsanity.com/paper/2106.01354/) (2021)
- [A survey on improving NLP models with human explanations](https://mlsanity.com/paper/2204.08892/) (2022)
- [A Fine-grained Interpretability Evaluation Benchmark for Neural NLP](https://mlsanity.com/paper/2205.11097/) (2022)
- [CLUES: A Benchmark for Learning Classifiers using Natural Language Explanations](https://mlsanity.com/paper/2204.07142/) (2022)
- [On Evaluating Explanation Utility for Human-AI Decision Making in NLP](https://mlsanity.com/paper/2407.03545/) (2024)
- [An Interpretability Evaluation Benchmark for Pre-trained Language Models](https://mlsanity.com/paper/2207.13948/) (2022)
- [A Logical Pattern Memory Pre-trained Model for Entailment Tree Generation](https://mlsanity.com/paper/2403.06410/) (2024)
- [Leveraging Multiple Teachers for Test-Time Adaptation of Language-Guided Classifiers](https://mlsanity.com/paper/2311.07538/) (2023)
- [AutoTriggER: Label-Efficient and Robust Named Entity Recognition with Auxiliary Trigger Extraction](https://mlsanity.com/paper/2109.04726/) (2021)
- [Explanation Graph Generation via Generative Pre-training over Synthetic Graphs](https://mlsanity.com/paper/2306.00652/) (2023)
- [A look under the hood of the Interactive Deep Learning Enterprise (No-IDLE)](https://mlsanity.com/paper/2406.19054/) (2024)
- [Teach me how to Label: Labeling Functions from Natural Language with Text-to-text Transformers](https://mlsanity.com/paper/2101.07138/) (2021)

## Abstract (arXiv)

Advances in machine reading comprehension (MRC) rely heavily on the collection of large scale human-annotated examples in the form of (question, paragraph, answer) triples. In contrast, humans are typically able to generalize with only a few examples, relying on deeper underlying world knowledge, linguistic sophistication, and/or simply superior deductive powers. In this paper, we focus on "teaching" machines reading comprehension, using a small number of semi-structured explanations that explicitly inform machines why answer spans are correct. We extract structured variables and rules from explanations and compose neural module teachers that annotate instances for training downstream MRC models. We use learnable neural modules and soft logic to handle linguistic variation and overcome sparse coverage; the modules are jointly optimized with the MRC model to improve final performance. On the SQuAD dataset, our proposed method achieves 70.14% F1 score with supervision from 26 explanations, comparable to plain supervised learning using 1,100 labeled instances, yielding a 12x speed up.