# Parameter-Efficient Detoxification with Contrastive Decoding

arXiv: https://arxiv.org/abs/2401.06947 · HTML: https://mlsanity.com/paper/2401.06947/

## Key facts

- Cited by 18 papers in the map (data through 14 Oct 2026).
- Builds on 30 of its 46 references that are in the map.
- First posted on arXiv 13 Jan 2024.

## Builds on (most cited first, 30 of 30)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Prefix-Tuning: Optimizing Continuous Prompts for Generation](https://mlsanity.com/paper/2101.00190/) (2021)
- [The Power of Scale for Parameter-Efficient Prompt Tuning](https://mlsanity.com/paper/2104.08691/) (2021)
- [The Curious Case of Neural Text Degeneration](https://mlsanity.com/paper/1904.09751/) (2019)
- [Don't Stop Pretraining: Adapt Language Models to Domains and Tasks](https://mlsanity.com/paper/2004.10964/) (2020)
- [RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Models](https://mlsanity.com/paper/2009.11462/) (2020)
- [A Diversity-Promoting Objective Function for Neural Conversation Models](https://mlsanity.com/paper/1510.03055/) (2015)
- [Few-Shot Parameter-Efficient Fine-Tuning is Better and Cheaper than In-Context Learning](https://mlsanity.com/paper/2205.05638/) (2022)
- [Plug and Play Language Models: A Simple Approach to Controlled Text Generation](https://mlsanity.com/paper/1912.02164/) (2019)
- [CTRL: A Conditional Transformer Language Model for Controllable Generation](https://mlsanity.com/paper/1909.05858/) (2019)
- [Universal Adversarial Triggers for Attacking and Analyzing NLP](https://mlsanity.com/paper/1908.07125/) (2019)
- [GPT Understands, Too](https://mlsanity.com/paper/2103.10385/) (2021)
- [Cross-Task Generalization via Natural Language Crowdsourcing Instructions](https://mlsanity.com/paper/2104.08773/) (2021)
- [Contrastive Decoding: Open-ended Text Generation as Optimization](https://mlsanity.com/paper/2210.15097/) (2022)
- [The Woman Worked as a Babysitter: On Biases in Language Generation](https://mlsanity.com/paper/1909.01326/) (2019)
- [Neural Text Generation with Unlikelihood Training](https://mlsanity.com/paper/1908.04319/) (2019)
- [DExperts: Decoding-Time Controlled Text Generation with Experts and Anti-Experts](https://mlsanity.com/paper/2105.03023/) (2021)
- [On Measuring Social Biases in Sentence Encoders](https://mlsanity.com/paper/1903.10561/) (2019)
- [GeDi: Generative Discriminator Guided Sequence Generation](https://mlsanity.com/paper/2009.06367/) (2020)
- [Do Prompt-Based Models Really Understand the Meaning of their Prompts?](https://mlsanity.com/paper/2109.01247/) (2021)
- [Measuring Bias in Contextualized Word Representations](https://mlsanity.com/paper/1906.07337/) (2019)
- [Gender Bias in Contextualized Word Embeddings](https://mlsanity.com/paper/1904.03310/) (2019)
- [A Token-level Reference-free Hallucination Detection Benchmark for Free-form Text Generation](https://mlsanity.com/paper/2104.08704/) (2021)
- [A Computational Approach to Politeness with Application to Social Factors](https://mlsanity.com/paper/1306.6078/) (2013)
- [Evaluating the Underlying Gender Bias in Contextualized Word Embeddings](https://mlsanity.com/paper/1904.08783/) (2019)
- [Transforming Delete, Retrieve, Generate Approach for Controlled Text Style Transfer](https://mlsanity.com/paper/1908.09368/) (2019)
- [Exploring the Limits of Domain-Adaptive Training for Detoxifying Large-Scale Language Models](https://mlsanity.com/paper/2202.04173/) (2022)
- [Polite Dialogue Generation Without Parallel Data](https://mlsanity.com/paper/1805.03162/) (2018)
- [Inference-Time Policy Adapters (IPA): Tailoring Extreme-Scale LMs without Fine-tuning](https://mlsanity.com/paper/2305.15065/) (2023)

## Built on by (most cited first, 18 of 18)

- [Merge, Ensemble, and Cooperate! A Survey on Collaborative Strategies in the Era of Large Language Models](https://mlsanity.com/paper/2407.06089/) (2024)
- [Linear Alignment: A Closed-form Solution for Aligning Human Preferences without Tuning and Feedback](https://mlsanity.com/paper/2401.11458/) (2024)
- [Model Surgery: Modulating LLM's Behavior Via Simple Parameter Editing](https://mlsanity.com/paper/2407.08770/) (2024)
- [Walking in Others' Shoes: How Perspective-Taking Guides Large Language Models in Reducing Toxicity and Bias](https://mlsanity.com/paper/2407.15366/) (2024)
- [Towards Inference-time Category-wise Safety Steering for Large Language Models](https://mlsanity.com/paper/2410.01174/) (2024)
- [On Almost Surely Safe Alignment of Large Language Models at Inference-Time](https://mlsanity.com/paper/2502.01208/) (2025)
- [SafeSteer: Interpretable Safety Steering with Refusal-Evasion in LLMs](https://mlsanity.com/paper/2506.04250/) (2025)
- [Unlocking Decoding-time Controllability: Gradient-Free Multi-Objective Alignment with Contrastive Prompts](https://mlsanity.com/paper/2408.05094/) (2024)
- [UniDetox: Universal Detoxification of Large Language Models via Dataset Distillation](https://mlsanity.com/paper/2504.20500/) (2025)
- [Toxicity in Online Platforms and AI Systems: A Survey of Needs, Challenges, Mitigations, and Future Directions](https://mlsanity.com/paper/2509.25539/) (2025)
- [Test-Time Detoxification without Training or Learning Anything](https://mlsanity.com/paper/2602.02498/) (2026)
- [IRepair: An Intent-Aware Approach to Repair Data-Driven Errors in Large Language Models](https://mlsanity.com/paper/2502.07072/) (2025)
- [Detoxification of Large Language Models through Output-layer Fusion with a Calibration Model](https://mlsanity.com/paper/2506.01266/) (2025)
- [Cleansing the Artificial Mind: A Self-Reflective Detoxification Framework for Large Language Models](https://mlsanity.com/paper/2601.11776/) (2026)
- [Fine-Grained Detoxification via Instance-Level Prefixes for Large Language Models](https://mlsanity.com/paper/2402.15202/) (2024)
- [Can LLMs Rank the Harmfulness of Smaller LLMs? We are Not There Yet](https://mlsanity.com/paper/2502.05291/) (2025)
- [Detoxification for LLM: From Dataset Itself](https://mlsanity.com/paper/2604.19124/) (2026)
- [Beyond Tokens: A Survey on Decoding Methods for Large Language and Vision-Language Models](https://mlsanity.com/paper/2608.14797/) (2026)

## Abstract (arXiv)

The field of natural language generation has witnessed significant advancements in recent years, including the development of controllable text generation techniques. However, controlling the attributes of the generated text remains a challenge, especially when aiming to avoid undesirable behavior such as toxicity. In this work, we introduce Detoxification Generator (DETOXIGEN), an inference-time algorithm that steers the generation away from unwanted styles. DETOXIGEN is an ensemble of a pre-trained language model (generator) and a detoxifier. The detoxifier is trained intentionally on the toxic data representative of the undesirable attribute, encouraging it to generate text in that style exclusively. During the actual generation, we use the trained detoxifier to produce undesirable tokens for the generator to contrast against at each decoding step. This approach directly informs the generator to avoid generating tokens that the detoxifier considers highly likely. We evaluate DETOXIGEN on the commonly used REALTOXICITYPROMPTS benchmark (Gehman et al., 2020) with various language models as generators. We find that it significantly outperforms previous approaches in detoxification metrics while not compromising on the generation quality. Moreover, the detoxifier is obtained by soft prompt-tuning using the same backbone language model as the generator. Hence, DETOXIGEN requires only a tiny amount of extra weights from the virtual tokens of the detoxifier to be loaded into GPU memory while decoding, making it a promising lightweight, practical, and parameter-efficient detoxification strategy.