用半监督学习提升大模型安全分类效果，减少对标注数据依赖。
Semi-Supervised Learning for Large Language Models Safety and Content Moderation
- 结合有标签和无标签数据，改进大模型安全分类器性能。
- 任务专用数据增强使安全检测准确率显著提升。
- 适合关注模型安全与内容审核的研究者与工程师。
大语言模型（LLMs）的安全性自其出现以来一直是研究重点，如今随着模型能力不断增强，这一问题更加突出。当前所有公开的LLM均设有若干安全防护机制，并已有多个用于训练安全分类器的数据集。然而，这些分类器的训练依赖大量标注数据，而这类数据获取困难、易出错，且常包含合成数据。为解决此问题，本文提出采用半监督学习方法，利用有标签与无标签数据共同训练，以提升安全任务表现。我们分析了该方法在提示（prompt）和模型响应两个层面的改进效果。此外，由于数据增强是半监督算法的核心，本文展示了任务专用增强的重要性——相比通用增强技术，其能显著提升性能。
原文摘要 · Abstract (English)
Safety for Large Language Models (LLMs) has been an ongoing research focus since their emergence and is even more relevant nowadays with the increasing capacity of those models. Currently, there are several guardrails in place for all public LLMs and multiple proposed datasets for training safety classifiers. However, training these safety classifiers relies on large quantities of labeled data, which can be problematic to acquire, prone to labeling errors, or often include synthetic data. To address these issues, we suggest a different approach: utilizing semi-supervised learning techniques, which leverage both labeled and unlabeled data, to improve the performance on the safety task. We analyze the improvements that these techniques can offer for both prompts given to Large Language Models and the responses to those requests. Moreover, since augmentation is the central part of semi-supervised algorithms, we demonstrate the importance of using task-specific augmentations, which significantly increase the performance when compared to general-purpose augmentation techniques.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。