# Finding the Needle in a Haystack: Unsupervised Rationale Extraction from Long Text Classifiers

arXiv: https://arxiv.org/abs/2303.07991 · HTML: https://mlsanity.com/paper/2303.07991/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 17 of its 21 references that are in the map.
- First posted on arXiv 14 Mar 2023.

## Builds on (most cited first, 17 of 17)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [Longformer: The Long-Document Transformer](https://mlsanity.com/paper/2004.05150/) (2020)
- [HuggingFace's Transformers: State-of-the-art Natural Language Processing](https://mlsanity.com/paper/1910.03771/) (2019)
- [Big Bird: Transformers for Longer Sequences](https://mlsanity.com/paper/2007.14062/) (2020)
- [What Does BERT Look At? An Analysis of BERT's Attention](https://mlsanity.com/paper/1906.04341/) (2019)
- [Are Sixteen Heads Really Better than One?](https://mlsanity.com/paper/1905.10650/) (2019)
- [ERASER: A Benchmark to Evaluate Rationalized NLP Models](https://mlsanity.com/paper/1911.03429/) (2019)
- [On the Stability of Fine-tuning BERT: Misconceptions, Explanations, and Strong Baselines](https://mlsanity.com/paper/2006.04884/) (2020)
- [Learning to Faithfully Rationalize by Construction](https://mlsanity.com/paper/2005.00115/) (2020)
- [Hierarchical Transformers for Long Document Classification](https://mlsanity.com/paper/1910.10781/) (2019)
- [Stop Explaining Black Box Machine Learning Models for High Stakes Decisions and Use Interpretable Models Instead](https://mlsanity.com/paper/1811.10154/) (2018)
- [Compositional Sequence Labeling Models for Error Detection in Learner Writing](https://mlsanity.com/paper/1607.06153/) (2016)
- [Weakly- and Semi-supervised Evidence Extraction](https://mlsanity.com/paper/2011.01459/) (2020)
- [Zero-shot Sequence Labeling: Transferring Knowledge from Sentences to Tokens](https://mlsanity.com/paper/1805.02214/) (2018)
- [Zero-shot Sequence Labeling for Transformer-based Sentence Classifiers](https://mlsanity.com/paper/2103.14465/) (2021)

## Built on by (most cited first, 1 of 1)

- [Prospector Heads: Generalized Feature Attribution for Large Models & Data](https://mlsanity.com/paper/2402.11729/) (2024)

## Abstract (arXiv)

Long-sequence transformers are designed to improve the representation of longer texts by language models and their performance on downstream document-level tasks. However, not much is understood about the quality of token-level predictions in long-form models. We investigate the performance of such architectures in the context of document classification with unsupervised rationale extraction. We find standard soft attention methods to perform significantly worse when combined with the Longformer language model. We propose a compositional soft attention architecture that applies RoBERTa sentence-wise to extract plausible rationales at the token-level. We find this method to significantly outperform Longformer-driven baselines on sentiment classification datasets, while also exhibiting significantly lower runtimes.