arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2001.08764cs.CL
收藏

减少由语言模型生成的非规范文本

Reducing Non-Normative Text Generation from Language Models

Georgia Institute of Technology(佐治亚理工学院)

Xiangyu Peng, Siyan Li, Spencer Frazier, Mark Riedl

摘要:

大规模的、基于transformer的语言模型(如GPT-2)是在从互联网上抓取的各种语料库上进行预训练的。因此,它们容易产生非规范性文本(即违反社会规范)。我们引入了一种微调GPT-2的技术,使用策略梯度强化学习技术和规范文本分类器来产生奖励和惩罚值。我们使用自动化和人类参与实验的五个数据集上评估我们的技术。当与标准和非标准生成文本的黄金标准人类判断相比时,标准文本分类器的准确率为81-90%。我们的规范性微调技术能够将非规范性文本减少27- 61%,具体取决于数据集。

英文摘要:

Large-scale, transformer-based language models such as GPT-2 are pretrained on diverse corpora scraped from the internet. Consequently, they are prone to generating non-normative text (i.e. in violation of social norms). We introduce a technique for fine-tuning GPT-2, using a policy gradient reinforcement learning technique and a normative text classifier to produce reward and punishment values. We evaluate our technique on five data sets using automated and human participant experiments. The normative text classifier is 81-90% accurate when compared to gold-standard human judgments of normative and non-normative generated text. Our normative fine-tuning technique is able to reduce non-normative text by 27-61%, depending on the data set.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑