arXivDaily arXiv每日学术速递 周一至周五更新
​通知:联邦节假日调休:哥伦布日 / 原住民日(Columbus Day / Indigenous Peoples' Day),今日(13日)无更新。
arXiv 2405.01490cs.CLcs.AI

指令微调时代的可控文本生成

Controllable Text Generation in the Instruction-Tuning Era

Carnegie Mellon University(卡内基梅隆大学);School of Computer Science(计算机科学学院)

Dhananjay Ashok, Barnabas Poczos

AI总结:

本文发布ConGenBench测试平台,发现提示方法在指令微调语言模型上优于传统可控生成方法,并提出自动生成约束数据集的算法以扩展可控性研究范围。

摘要:

虽然大多数关于可控文本生成的研究都集中在引导基座语言模型上,但新兴的指令微调和提示范式为可控性提供了一种替代方法。我们汇编并发布了ConGenBench,这是一个包含17个不同可控生成任务的测试平台,并使用其中的一个子集来基准测试9种不同基线和方方法在指令微调语言模型上的性能。令我们惊讶的是,我们发现基于提示的方法在大多数数据集和任务上优于可控文本生成方法,这凸显了专门针对指令微调语言模型的可控文本生成研究的必要性。基于提示的方法在大多数风格任务上与人类表现相当,但在结构任务上表现滞后,这突出表明需要研究更多样化的约束和更具挑战性的风格任务。为了促进此类研究,我们提供了一种算法,该算法仅使用任务数据集和具有上下文能力的大型语言模型来自动生成约束数据集。该方法消除了该领域对预先整理的约束数据集的依赖,从而极大地扩展了未来可以研究的约束范围。

英文摘要:

While most research on controllable text generation has focused on steering base Language Models, the emerging instruction-tuning and prompting paradigm offers an alternate approach to controllability. We compile and release ConGenBench, a testbed of 17 different controllable generation tasks, using a subset of it to benchmark the performance of 9 different baselines and methods on Instruction-tuned Language Models. To our surprise, we find that prompting-based approaches outperform controllable text generation methods on most datasets and tasks, highlighting a need for research on controllable text generation with Instruction-tuned Language Models in specific. Prompt-based approaches match human performance on most stylistic tasks while lagging on structural tasks, foregrounding a need to study more varied constraints and more challenging stylistic tasks. To facilitate such research, we provide an algorithm that uses only a task dataset and a Large Language Model with in-context capabilities to automatically generate a constraint dataset. This method eliminates the fields dependence on pre-curated constraint datasets, hence vastly expanding the range of constraints that can be studied in the future.

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。

↑