# Automatically Neutralizing Subjective Bias in Text

arXiv: https://arxiv.org/abs/1911.09709 · HTML: https://mlsanity.com/paper/1911.09709/

## Key facts

- Cited by 105 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 21 Nov 2019.

## Built on by (most cited first, 30 of 105)

- [Bias and Fairness in Large Language Models: A Survey](https://mlsanity.com/paper/2309.00770/) (2023)
- [Societal Biases in Language Generation: Progress and Challenges](https://mlsanity.com/paper/2105.04054/) (2021)
- [A Survey of Controllable Text Generation using Transformer-based Pre-trained Language Models](https://mlsanity.com/paper/2201.05337/) (2022)
- [Symbolic Learning Enables Self-Evolving Agents](https://mlsanity.com/paper/2406.18532/) (2024)
- [Navigating the Grey Area: How Expressions of Uncertainty and Overconfidence Affect Language Models](https://mlsanity.com/paper/2302.13439/) (2023)
- [Controlled Text Generation with Natural Language Instructions](https://mlsanity.com/paper/2304.14293/) (2023)
- [PEER: A Collaborative Language Model](https://mlsanity.com/paper/2208.11663/) (2022)
- [Language Generation Models Can Cause Harm: So What Can We Do About It? An Actionable Survey](https://mlsanity.com/paper/2210.07700/) (2022)
- [Towards Facilitating Empathic Conversations in Online Mental Health Support: A Reinforcement Learning Approach](https://mlsanity.com/paper/2101.07714/) (2021)
- [CoEdIT: Text Editing by Task-Specific Instruction Tuning](https://mlsanity.com/paper/2305.09857/) (2023)
- [RewriteLM: An Instruction-Tuned Large Language Model for Text Rewriting](https://mlsanity.com/paper/2305.15685/) (2023)
- [Do LLMs Understand Social Knowledge? Evaluating the Sociability of Large Language Models with SocKET Benchmark](https://mlsanity.com/paper/2305.14938/) (2023)
- [LENS: A Learnable Evaluation Metric for Text Simplification](https://mlsanity.com/paper/2212.09739/) (2022)
- [Can AI writing be salvaged? Mitigating Idiosyncrasies and Improving Human-AI Alignment in the Writing Process through Edits](https://mlsanity.com/paper/2409.14509/) (2024)
- [Are Personalized Stochastic Parrots More Dangerous? Evaluating Persona Biases in Dialogue Systems](https://mlsanity.com/paper/2310.05280/) (2023)
- [Beyond the Chat: Executable and Verifiable Text-Editing with LLMs](https://mlsanity.com/paper/2309.15337/) (2023)
- [Multi-Task Instruction Tuning of LLaMa for Specific Scenarios: A Preliminary Study on Writing Assistance](https://mlsanity.com/paper/2305.13225/) (2023)
- [Civil Rephrases Of Toxic Texts With Self-Supervised Transformers](https://mlsanity.com/paper/2102.05456/) (2021)
- [Inducing Positive Perspectives with Text Reframing](https://mlsanity.com/paper/2204.02952/) (2022)
- [EditEval: An Instruction-Based Benchmark for Text Improvements](https://mlsanity.com/paper/2209.13331/) (2022)
- [Learning Interpretable Style Embeddings via Prompting LLMs](https://mlsanity.com/paper/2305.12696/) (2023)
- [NEWTS: A Corpus for News Topic-Focused Summarization](https://mlsanity.com/paper/2205.15661/) (2022)
- [arXivEdits: Understanding the Human Revision Process in Scientific Writing](https://mlsanity.com/paper/2210.15067/) (2022)
- [Detect and Perturb: Neutral Rewriting of Biased and Sensitive Text via Gradient-based Decoding](https://mlsanity.com/paper/2109.11708/) (2021)
- [NBIAS: A Natural Language Processing Framework for Bias Identification in Text](https://mlsanity.com/paper/2308.01681/) (2023)
- [Gender Biases and Where to Find Them: Exploring Gender Bias in Pre-Trained Transformer-based Language Models Using Movement Pruning](https://mlsanity.com/paper/2207.02463/) (2022)
- [Fairness and Bias in Multimodal AI: A Survey](https://mlsanity.com/paper/2406.19097/) (2024)
- [REVOLVE: Optimizing AI Systems by Tracking Response Evolution in Textual Optimization](https://mlsanity.com/paper/2412.03092/) (2024)
- [A Review of Human Evaluation for Style Transfer](https://mlsanity.com/paper/2106.04747/) (2021)
- [Introducing MBIB -- the first Media Bias Identification Benchmark Task and Dataset Collection](https://mlsanity.com/paper/2304.13148/) (2023)

## Abstract (arXiv)

Texts like news, encyclopedias, and some social media strive for objectivity. Yet bias in the form of inappropriate subjectivity - introducing attitudes via framing, presupposing truth, and casting doubt - remains ubiquitous. This kind of bias erodes our collective trust and fuels social conflict. To address this issue, we introduce a novel testbed for natural language generation: automatically bringing inappropriately subjective text into a neutral point of view ("neutralizing" biased text). We also offer the first parallel corpus of biased language. The corpus contains 180,000 sentence pairs and originates from Wikipedia edits that removed various framings, presuppositions, and attitudes from biased sentences. Last, we propose two strong encoder-decoder baselines for the task. A straightforward yet opaque CONCURRENT system uses a BERT encoder to identify subjective words as part of the generation process. An interpretable and controllable MODULAR algorithm separates these steps, using (1) a BERT-based classifier to identify problematic words and (2) a novel join embedding through which the classifier can edit the hidden states of the encoder. Large-scale human evaluation across four domains (encyclopedias, news headlines, books, and political speeches) suggests that these algorithms are a first step towards the automatic identification and reduction of bias.