Out of Order: How important is the sequential order of words in a sentence in Natural Language Understanding tasks?

By Thang M. Pham, Trung Bui, Long Mai, Anh Totti Nguyen. 2020

129 citations recorded by the source corpus.

Do state-of-the-art natural language understanding models care about word order - one of the most important characteristics of a sequence? Not always! We found 75% to 90% of the correct predictions of BERT-based classifiers, trained on many GLUE tasks, remain constant after input words are randomly shuffled. Despite BERT embeddings are famously contextual, the contribution of each individual word to downstream tasks is almost unchanged even after the word's context is shuffled. BERT-based models are able to exploit superficial cues (e.g. the sentiment of keywords in sentiment analysis; or the word-wise similarity between sequence-pair inputs in natural language inference) to make correct decisions when tokens are arranged in random orders. Encouraging classifiers to capture word order information improves the performance on most GLUE tasks, SQuAD 2.0 and out-of-samples. Our work suggests that many GLUE tasks are not challenging machines to understand the meaning of a sentence.

Subfields: Artificial Intelligence, Natural Language Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Natural Language Processing

Citations in the corpus

Can language models learn from explanations in context? (2022, background): “First, we ensured that the beneﬁts of explanations are not due simply to word-level features, as smaller unsupervised text models have been found to ignore word order in some contexts (e.g., Pham et al., 2020; O’Connor and Andreas, 2021, but cf. Pa-padimitriou et al., 2022).”
Do Prompt-Based Models Really Understand the Meaning of Their Prompts? (2021, positive): “Our main argument throughout the paper shares the same logic as a recent line of studies (Sinha et al., 2021; O’Connor and Andreas, 2021; Pham et al., 2021; Gupta et al., 2021) which argue that the fact that LMs achieve good performance under ideal conditions is insufﬁcient to establish language…”
What Context Features Can Transformer Language Models Use? (2021, methodology): “(1) and (2) were also recently explored by Pham et al. (2020) in models for entailment, and more complex shufﬂing procedures have been explored in neuroscience contexts (Mollica et al., 2020).”
Rissanen Data Analysis: Examining Dataset Characteristics via Description Length (2021, methodology): “view word order (Pham et al., 2020; Sinha et al., 2020; Gupta et al., 2021) or words of different types (Sugawara et al.”
When and why vision-language models behave like bags-of-words, and what to do about it? (2022, background): “Order information in language and vision: Several works have highlighted the lack of word order sensitivity in large language models (Hessel & Schoﬁeld, 2021; O’Connor & Andreas, 2021; Pham et al., 2021; Sinha et al., 2021).”
Reasoning with Transformer-based Models: Deep Learning, but Shallow Reasoning (2021, background): “Diﬀerent studies [Ettinger, 2020, Sankar et al., 2019, Pham et al., 2020, Gupta et al., 2021] have shown that BERT-like models are unperturbed by grammatically incorrect sentences: If presented with a sentence of randomly shuﬄed words, they will still reply correctly.”
When classifying arguments, BERT doesn’t care about word order...except when it matters (2022, background): “…a spate of recent work has shown that they are surprisingly invariant to scrambling word order (Sinha et al., 2021; Hessel and Schofield, 2021; Pham et al., 2020; Gupta et al., 2021; O’Connor and Andreas, 2021) and that grammatical knowledge like part of speech, often attributed to…”
Masked Language Modeling and the Distributional Hypothesis: Order Word Matters Pre-training for Little (2021, methodology): “Sensitivity to word order for NLU tasks has recently been inves- tigated concurrently by Sinha et al. (2020), Pham et al. (2020), and Gupta et al. (2021).”
The Case for Translation-Invariant Self-Attention in Transformer-Based Language Models (2021, methodology): “This agrees with experiments shuffling the order of words (Pham et al., 2020; Sinha et al., 2021) finding that modern language models tend to focus mainly on higherorder word co-occurrences, rather than word order, and suggests that word-order information is underutilized in state-of-the-art languag”
Sometimes We Want Ungrammatical Translations (2021, unknown): “(Si et al., 2019; Sinha et al., 2020; Pham et al., 2020; Gupta et al., 2021; Sinha et al., 2021).”