arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2110.04541cs.CLcs.LG
收藏

上下文学习的归纳偏差:对预训练样例设计的再思考

The Inductive Bias of In-Context Learning: Rethinking Pretraining Example Design

The Hebrew University of Jerusalem(耶路撒冷希伯来大学)

Yoav Levine, Noam Wies, Daniel Jannai, Dan Navon, Yedid Hoshen, Amnon Shashua

摘要:

在大型语料库上预训练神经语言模型(NLM)涉及将文本分块成训练样本,这些样本是神经架构可处理的大小的连续文本段。我们强调了这种常见做法引入的偏见:我们证明了预训练的NLM可以对出现在同一训练示例中的文本段之间的依赖关系进行建模,而不是出现在不同训练样本中的文本段之间的依赖关系。这个直观的结果具有双重作用。首先,它形式化了最近成功的NLM培训策略背后的动机,这些策略是针对预训练和微调阶段提出的,乍一看并不一定相关。其次,我们的结果清楚地表明,NLM预训练的进一步改进,有利于自然语言理解任务。作为一个例子,我们提出了“kNN预训练”:我们表明,包括语义相关的非相邻句子在同一预训练的例子产生改进的句子表示和开放域问答能力。这种理论上的自由度为预训练的例子设计表明新的培训计划,自我改进的表示。

英文摘要:

Pretraining Neural Language Models (NLMs) over a large corpus involves chunking the text into training examples, which are contiguous text segments of sizes processable by the neural architecture. We highlight a bias introduced by this common practice: we prove that the pretrained NLM can model much stronger dependencies between text segments that appeared in the same training example, than it can between text segments that appeared in different training examples. This intuitive result has a twofold role. First, it formalizes the motivation behind a broad line of recent successful NLM training heuristics, proposed for the pretraining and fine-tuning stages, which do not necessarily appear related at first glance. Second, our result clearly indicates further improvements to be made in NLM pretraining for the benefit of Natural Language Understanding tasks. As an example, we propose "kNN-Pretraining": we show that including semantically related non-neighboring sentences in the same pretraining example yields improved sentence representations and open domain question answering abilities. This theoretically motivated degree of freedom for pretraining example design indicates new training schemes for self-improving representations.

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。

↑