paper-with-me

Papers

Contrastive Perplexity for Controlled Generation: An Application in Detoxifying Large Language Models

2024-01-16 · Tassilo Klein, Moin Nabi

The generation of undesirable and factually incorrect content of large language models poses a significant challenge and remains largely an unsolved issue. This paper studies the integration of a contrastive learning objective for fine-tuning LLMs for implicit knowledge editing and controlled text generation. Optimizing the training objective entails aligning text perplexities in a contrastive fashion. To facilitate training the model in a self-supervised fashion, we leverage an off-the-shelf LLM for training data generation. We showcase applicability in the domain of detoxification. Herein, the proposed approach leads to a significant decrease in the generation of toxic content while preserving general utility for downstream tasks such as commonsense reasoning and reading comprehension. The proposed approach is conceptually simple but empirically powerful.

📄 PDF Abstract BibTeX arXiv:2401.08491

Code (0)

등록된 구현이 없습니다.

Tasks

Contrastive Learningknowledge editingReading ComprehensionText Generation

Methods 이 논문이 사용한 방법론

Contrastive Learning 설명 없음

Similar Papers 제목 키워드 기반

CFL: Causally Fair Language Models Through Token-level Attribute Controlled Generation

2023-06-01 · Rahul Madhavan, Rishabh Garg, Kahini Wadhawan, Sameep Mehta

We propose a method to control the attributes of Language Models (LMs) for the text generation task using Causal Average Treatment Effect (ATE) scores and counterfactual augmentation. We explore this method, in the conte…

AttributecounterfactualText Generation

Exploring the Limits of Domain-Adaptive Training for Detoxifying Large-Scale Language Models

2022-02-08 · Boxin Wang, Wei Ping, Chaowei Xiao, Peng Xu 외

Pre-trained language models (LMs) are shown to easily generate toxic language. In this work, we systematically explore domain-adaptive training to reduce the toxicity of language models. We conduct this study on three di…

Unified Detoxifying and Debiasing in Language Generation via Inference-time Adaptive Optimization

2022-10-10 · Zonghan Yang, Xiaoyuan Yi, Peng Li, Yang Liu 외

Warning: this paper contains model outputs exhibiting offensiveness and biases. Recently pre-trained language models (PLMs) have prospered in various natural language generation (NLG) tasks due to their ability to genera…

Text Generation

UniDetox: Universal Detoxification of Large Language Models via Dataset Distillation

2025-04-29 · Huimin Lu, Masaru Isonuma, Junichiro Mori, Ichiro Sakata

We present UniDetox, a universally applicable method designed to mitigate toxicity across various large language models (LLMs). Previous detoxification methods are typically model-specific, addressing only individual mod…

Dataset DistillationLanguage ModelingLanguage Modelling

Multi-Attribute Controlled Text Generation with Contrastive-Generator and External-Discriminator

2022-10-01 · COLING 2022 10 · Guisheng Liu, Yi Li, Yanqing Guo, Xiangyang Luo 외

Though existing researches have achieved impressive results in controlled text generation, they focus mainly on single-attribute control. However, in applications like automatic comments, the topic and sentiment need to …

AttributeComment GenerationText Generation