Civil Rephrases Of Toxic Texts With Self-Supervised Transformers (2021.eacl-main)
Copied to clipboard
| Challenge: | et al., 2018a): a poor phrasing may make the conversation go awry. |
| Approach: | They propose a model that can help suggest rephrasings of toxic comments in a more civil manner. |
| Outcome: | The proposed model generates sentences that are more fluent and better at preserving the initial content compared to earlier systems and human evaluation. |
Similar Papers
ModelCitizens: Representing Community Voices in Online Safety (2025.emnlp-main)
Copied to clipboard
Ashima Suvarna, Christina A Chance, Karolina Naranjo, Hamid Palangi, Sophie Hao, Thomas Hartvigsen, Saadia Gabriel
| Challenge: | Existing toxic language detection models are trained on annotations that collapse diverse perspectives into a single ground truth. |
| Approach: | They propose to augment social media posts with conversational scenarios to reflect the impact of conversational context on toxicity. |
| Outcome: | The proposed model outperforms existing models on social media with conversational scenarios. |
COUNT: COntrastive UNlikelihood Text Style Transfer for Text Detoxification (2023.findings-emnlp)
Copied to clipboard
Mohammad Mahdi Abdollah Pour, Parsa Farinneya, Manasa Bharadwaj, Nikhil Verma, Ali Pesaranghader, Scott Sanner
| Challenge: | Text detoxification is a task to ensure the generation of non-toxic and safe text. |
| Approach: | They propose a novel contrastive unlikelihood objective that combines rephrasing and identity mapping to effectively isolate and focus learning on non-toxic style transfer. |
| Outcome: | The proposed method achieves significant improvements in fluency, content preservation, and detoxification on two parallel datasets. |
RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Models (2020.findings-emnlp)
Copied to clipboard
| Challenge: | Pretrained neural language models (LMs) are prone to generating toxic language . we investigate the extent to which pretrained LMs can be prompted to generate toxic language. |
| Approach: | They propose to use a dataset to test the extent to which pretrained LMs can generate toxic language . they use toxicity-based text generation algorithms to generate toxic text from natural prompts . |
| Outcome: | The proposed model can degenerate into toxic text even from seemingly innocuous prompts. |
Text Detoxification: Data Efficiency, Semantic Preservation and Model Generalization (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods for detoxification of text often rely on manually annotated data . xiangli: "detoxification of texts is a powerful way to remove toxic content" |
| Approach: | They propose a reinforcement learning framework that optimizes detoxification and semantic preservation without annotating large amounts of data. |
| Outcome: | The proposed method overcomes major limitations and surpasses humanannotated references across multiple benchmarks. |
Text Detoxification using Large Pre-trained Neural Models (2021.emnlp-main)
Copied to clipboard
David Dale, Anton Voronov, Daryna Dementieva, Varvara Logacheva, Olga Kozlova, Nikita Semenov, Alexander Panchenko
| Challenge: | Existing studies on text detoxification cast this task as style transfer . text detox requires better preservation of the original meaning, authors argue . |
| Approach: | They propose two unsupervised methods for eliminating toxicity in text . they use a paraphraser guided by style-trained language models to keep the text content . |
| Outcome: | The proposed methods yield new SOTA results. |
From the Detection of Toxic Spans in Online Discussions to the Analysis of Toxic-to-Civil Transfer (2022.acl-long)
Copied to clipboard
| Challenge: | a dataset of English posts with annotations of toxic spans is released . sequence labeling models perform best, but rationale extraction methods are promising . |
| Approach: | They propose a dataset for toxic spans detection that includes an annotation of toxic posts . they propose to add generic rationale extraction mechanisms to the model to obtain toxic span information . |
| Outcome: | The proposed framework is based on a dataset of English posts with toxic span annotations . it shows that sequence labeling models perform best, but that rationale extraction methods are promising . |
Contrastive Perplexity for Controlled Generation: An Application in Detoxifying Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to generate toxic content by large language models are based on pipelines . current approaches focus on preserving performance while effectively mitigating toxicity . |
| Approach: | They propose a framework for implicit knowledge editing and controlled text generation by using hard negatives. |
| Outcome: | The proposed framework significantly reduces toxic generation while maintaining strong performance on downstream tasks. |
Challenges in Automated Debiasing for Toxic Language Detection (2021.eacl-main)
Copied to clipboard
| Challenge: | Existing methods for debiasing toxic language data are limited in their ability to prevent biased behavior in toxic language detection systems. |
| Approach: | They propose to debiase toxic language detection models using lexical and dialectal markers using synthetic labels instead of traditional methods. |
| Outcome: | The proposed method reduces dialectal associations with toxicity despite the use of synthetic labels . |
Breaking Bad Tokens: Detoxification of LLMs Using Sparse Autoencoders (2025.emnlp-main)
Copied to clipboard
| Challenge: | Large language models (LLMs) are ubiquitous in user-facing applications, yet they still generate undesirable toxic outputs, including profanity, vulgarity, and derogatory remarks. |
| Approach: | They leverage sparse autoencoders to identify toxicity-related directions in residual stream of large language models and perform targeted activation steering using the corresponding decoder vectors. |
| Outcome: | The proposed models surpass baselines in reducing toxicity by up to 20%, though fluency can degrade noticeably on GPT-2 Small and Gemma-2-2B. |
Detoxifying Text with MaRCo: Controllable Revision with Experts and Anti-Experts (2023.acl-short)
Copied to clipboard
| Challenge: | Text detoxification can mitigate the harms of toxicity by rephrasing text to remove offensive meaning, but subtle toxicity remains challenging to tackle. |
| Approach: | They propose a text detoxification algorithm that combines controllable generation and text rewriting methods using a Product of Experts and autoencoder language models to find candidate words to mask and potentially replace. |
| Outcome: | The proposed method outperforms baselines on automatic metrics and is preferred 2.1 times more in human evaluation. |