arXiv cs.CL· Esmeralda S. Whitammer, Sameera Lanka, Pranav Goel, Nebojsa Jojic·· 11 小时前AI 评分26
研究：通过迭代打乱（IBIS）检验词序是否编码语义
Studying word order through iterative shuffling
AI 导读
研究提出 IBIS（inference by iterative shuffling），一种无需额外训练、可作用于任意黑盒语言模型的高效流程，用于找出词袋在固定语言模型下似然最高的排列，效果优于现有词序算法。
正文 · 原文
Abstract:As neural language models approach human performance on NLP benchmark tasks, their advances are widely seen as evidence of an increasingly complex understanding of syntax. This view rests upon a hypothesis that has not yet been empirically tested: that word order encodes meaning essential to performing these tasks. We refute this hypothesis in many cases: in the GLUE suite and in various genres of English text, the words in a sentence or phrase can rarely be permuted to form a phrase carrying substantially different information. Our surprising result relies on inference by iterative shuffling (IBIS), a novel, efficient procedure that finds the ordering of a bag of words having the highest likelihood under a fixed language model. IBIS can use any black-box model without additional training and is superior to existing word ordering algorithms. Coalescing our findings, we discuss how shuffling inference procedures such as IBIS can benefit language modeling and constrained generation.
| Comments: | EMNLP 2021; code: this https URL |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2109.04867 [cs.CL] |
| (or arXiv:2109.04867v2 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2109.04867 arXiv-issued DOI via DataCite |
|
| Related DOI: | https://doi.org/10.18653/v1/2021.emnlp-main.809
DOI(s) linking to related resources |
Submission history
From: Esmeralda Whitammer [view email]
[v1]
Fri, 10 Sep 2021 13:27:06 UTC (6,467 KB)
[v2]
Thu, 5 Nov 2026 12:23:23 UTC (6,463 KB)
来源：arXiv cs.CL · arxiv.org