arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2302.09185cs.CLcs.AIcs.LG
收藏

开放文本生成中大语言模型的提示约束能力

Bounding the Capabilities of Large Language Models in Open Text Generation with Prompt Constraints

Georgia Institute of Technology(佐治亚理工学院);Shanghai Jiao Tong University(上海交通大学);Google(谷歌);Stanford University(斯坦福大学)

Albert Lu, Hongxin Zhang, Yanzhe Zhang, Xuezhi Wang, Diyi Yang

摘要:

开放式生成模型的局限性尚不清楚,但越来越重要。是什么让他们成功,又是什么让他们失败?在本文中,我们采取了一种以知识为中心的方法来分析和界定开放式生成模型的能力。我们提出了一个通用的分析方法,两个具有挑战性的提示约束类型:结构和风格。这些约束类型被分类为一组定义良好的约束,这些约束可通过单个提示进行分析。然后,我们系统地创建一组简单,自然和有用的提示,以强大地分析每个单独的约束。使用GPT-3 text-davinci-002模型作为案例研究,我们从提示集合中生成输出并分析模型的生成失败。我们还展示了我们所提出的方法在其他大型模型上的推广性,如BLOOM和OPT. We的结果和我们的上下文缓解策略揭示了未来研究面临的挑战。我们已在https://github.com/SALT-NLP/Bound-Cap-LLM上公开发布了我们的代码。

英文摘要:

The limits of open-ended generative models are unclear, yet increasingly important. What causes them to succeed and what causes them to fail? In this paper, we take a prompt-centric approach to analyzing and bounding the abilities of open-ended generative models. We present a generic methodology of analysis with two challenging prompt constraint types: structural and stylistic. These constraint types are categorized into a set of well-defined constraints that are analyzable by a single prompt. We then systematically create a diverse set of simple, natural, and useful prompts to robustly analyze each individual constraint. Using the GPT-3 text-davinci-002 model as a case study, we generate outputs from our collection of prompts and analyze the model's generative failures. We also show the generalizability of our proposed method on other large models like BLOOM and OPT. Our results and our in-context mitigation strategies reveal open challenges for future research. We have publicly released our code at https://github.com/SALT-NLP/Bound-Cap-LLM.

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。

↑