arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2107.09285cs.CLcs.AI
收藏

神经抽象：支持构建的抽象语言学习

Neural Abstructions: Abstractions that Support Construction for Grounded Language Learning

Stanford University（斯坦福大学）

Kaylee Burns, Christopher D. Manning, Li Fei-Fei

备注：
17 pages, 10 figures
阅读 PDF
arXiv 页面 ↗HTML 原文 ↗

摘要：

虽然虚拟代理越来越多地位于环境中，自然语言是最有效的模式与人类的互动，这些交流很少被用作学习的机会。有效地利用语言交互需要解决两种最常见的语言基础方法的局限性：建立在固定对象类别之上的语义解析器精确但不灵活，端到端模型表达能力最强，但变化无常且不透明。我们的目标是开发一个系统，平衡每种方法的优势，使用户可以教代理新的指令，概括从一个单一的例子。我们介绍了神经abstructions的想法：一组约束的推理过程中的标签条件生成模型，可以影响的意义上的标签在上下文中。从对抽象操作的核心程序设计语言开始，用户可以定义从自然语言到动作的越来越复杂的映射。我们表明，使用这种方法，用户群体能够在Minecraft中为开放式房屋修改任务构建语义解析器。语义解析器的结果是灵活的和表达：来自重新定义的话语的百分比稳步增加的过程中，191个总的交流，达到28%的最终值。

英文摘要：

Although virtual agents are increasingly situated in environments where natural language is the most effective mode of interaction with humans, these exchanges are rarely used as an opportunity for learning. Leveraging language interactions effectively requires addressing limitations in the two most common approaches to language grounding: semantic parsers built on top of fixed object categories are precise but inflexible and end-to-end models are maximally expressive, but fickle and opaque. Our goal is to develop a system that balances the strengths of each approach so that users can teach agents new instructions that generalize broadly from a single example. We introduce the idea of neural abstructions: a set of constraints on the inference procedure of a label-conditioned generative model that can affect the meaning of the label in context. Starting from a core programming language that operates over abstructions, users can define increasingly complex mappings from natural language to actions. We show that with this method a user population is able to build a semantic parser for an open-ended house modification task in Minecraft. The semantic parser that results is both flexible and expressive: the percentage of utterances sourced from redefinitions increases steadily over the course of 191 total exchanges, achieving a final value of 28%.

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。

↑