CoChat

Do NLP Models Know Numbers? Probing Numeracy in Embeddings

Open access
Wallace, Eric, Wang, Yizhong, Li, Sujian, Singh, Sameer, Gardner, Matt · 2019 · Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)

The DOI matches its Crossref record; 29 of the 29 references it cites were found in a registry or an index.

Abstract

Eric Wallace, Yizhong Wang, Sujian Li, Sameer Singh, Matt Gardner. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 2019.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.18653/v1/d19-1534

Cites

29 of 29 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 We train a probing model to decode a number from its word embedding over a random 80% of the integers from [-500, 500], e.g., '71' → 71.0. We plot the model's predictions for all numbers from [-2000, 2000]. The model accurately decodes numbers within the training range (in blue), i.e., pre-trained embeddings like GloVe and BERT capture numeracy. However, the probe fails to extrapolate to larger numbers (in red). The Char-CNN (e) and Char-LSTM (f) are trained jointly with the probing model.
Figure 1 We train a probing model to decode a number from its word embedding over a random 80% of the integers from [-500, 500], e.g., '71' → 71.0. We plot the model's predictions for all numbers from [-2000, 2000]. The model accurately decodes numbers within the training range (in blue), i.e., pre-trained embeddings like GloVe and BERT capture numeracy. However, the probe fails to extrapolate to larger numbers (in red). The Char-CNN (e) and Char-LSTM (f) are trained jointly with the probing model.CC-BY · From the PDF
Figure 3 Our probing setup. We pass numbers through a pre-trained embedder (e.g., BERT, GloVe) and train a probing model to solve numerical tasks such as finding a list's maximum, decoding a number, or adding two numbers. If the probing model generalizes to held-out numbers, the pre-trained embeddings must contain numerical information. We provide numbers as either words (shown here), digits ('9'), floats ('9.1'), or negatives ('-9').
Figure 3 Our probing setup. We pass numbers through a pre-trained embedder (e.g., BERT, GloVe) and train a probing model to solve numerical tasks such as finding a list's maximum, decoding a number, or adding two numbers. If the probing model generalizes to held-out numbers, the pre-trained embeddings must contain numerical information. We provide numbers as either words (shown here), digits ('9'), floats ('9.1'), or negatives ('-9').CC-BY · From the PDF

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

  • BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

    Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2018) · arXiv · doi:10.48550/arxiv.1810.04805 · Cited by 73,173 · cited 3×

    “, 2018), BERT (Devlin et al., 2019), and GloVe (Pennington et al.”
    Background
  • BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

    Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2019) · Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers) · doi:10.18653/v1/n19-1423 · Cited by 54,325 · cited 3×

    “, 2018), BERT (Devlin et al., 2019), and GloVe (Pennington et al.”
    Background
  • Glove: Global Vectors for Word Representation

    Pennington, J., Socher, R., & Manning, C. (2014) · Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.3115/v1/d14-1162 · Cited by 34,071 · cited 4×

    “We find that all widely-used pre-trained embed-dings, e.g., ELMo (Peters et al., 2018), BERT (De-vlin et al., 2019), and GloVe (Pennington et al., 2014), capture numeracy: number magnitude is present in the embeddings, even for numbers in the thousands.”
    BackgroundMethodInfluential
  • Deep Contextualized Word Representations

    Peters, M., Neumann, M., Iyyer, M., Gardner, M., Clark, C., Lee, K., & Zettlemoyer, L. (2018) · Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Papers) · doi:10.18653/v1/n18-1202 · Cited by 11,983 · cited 3×

    “We find that all widely-used pre-trained embed-dings, e.g., ELMo (Peters et al., 2018), BERT (De-vlin et al., 2019), and GloVe (Pennington et al., 2014), capture numeracy: number magnitude is present in the embeddings, even for numbers in the thousands.”
    MethodSelf-citation
  • SQuAD: 100,000+ Questions for Machine Comprehension of Text

    Rajpurkar, P., Zhang, J., Lopyrev, K., & Liang, P. (2016) · Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/d16-1264 · Cited by 6,208 · cited 2×

    “The dataset’s input-output format is a superset of SQuAD (Rajpurkar et al., 2016): the an... JaMarcus Russell completed a 91-yard touchdown pass to rookie wide receiver Chaz Schilens .”
    Method
  • Understanding Black-box Predictions via Influence Functions

    Koh, P. W., & Liang, P. (2017) · arXiv · doi:10.48550/arxiv.1703.04730 · Cited by 1,514 · cited 2×

    “, influence functions (Koh and Liang, 2017; Brunet et al., 2019)).”
    Background
  • Bidirectional Attention Flow for Machine Comprehension

    Seo, M. J., Kembhavi, A., Farhadi, A., & Hajishirzi, H. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1611.01603 · Cited by 1,445 · cited 2×

    “As in typical span selection models (Seo et al., 2017), an LSTM reads the list of token embeddings, and a weight matrix and softmax function assign a probability to each index using the model’s hidden state.”
    BackgroundMethod
  • Neural Turing Machines

    Graves, A., Wayne, G., & Danihelka, I. (2014) · arXiv · doi:10.48550/arxiv.1410.5401 · Cited by 1,058 · cited 2×

    “Synthetic Numerical Tasks Similar to our synthetic numerical reasoning tasks, other work considers sorting (Graves et al., 2014), counting (Weiss et al., 2018), or decoding tasks (Trask et al., 2018).”
    Method
  • Assessing the Ability of LSTMs to Learn Syntax-Sensitive Dependencies

    Linzen, T., Dupoux, E., & Goldberg, Y. (2016) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl_a_00115 · Cited by 737 · cited 1×

    “Khandel-wal et al. (2018) show how language models leverage context, while Linzen et al. (2016) demonstrate that language models understand subject-verb agreement.”
    Background
  • Linguistic Knowledge and Transferability of Contextual Representations

    Liu, N. F., Gardner, M., Belinkov, Y., Peters, M. E., & Smith, N. A. (2019) · Proceedings of the 2019 Conference of the North · doi:10.18653/v1/n19-1112 · Cited by 621 · cited 2×

    “Probing Models Our probes of numeracy parallel work in understanding the linguistic capabilities (literacy) of neural models (Conneau et al., 2018; Liu et al., 2019).”
    BackgroundSelf-citation

Show all 29 references

Research works printed but not linked: 5

  1. [1]1. Yossi Adi, Einat Kermany, Yonatan Belinkov, Ofer Lavi, and Yoav Goldberg. 2017. Finegrained anal- ysis of sentence embeddings using auxiliary predic- tion tasks. In ICLR.
  2. [13]13. Nelson F Liu, Matt Gardner, Yonatan Belinkov, Matthew Peters, and Noah A Smith. 2019. Linguis- tic knowledge and transferability of contextual rep- resentations. In NAACL.
  3. [14]14. Tomas Mikolov, Edouard Grave, Piotr Bojanowski, Christian Puhrsch, and Armand Joulin. 2018. Ad- vances in pre-training distributed word representa- tions. In LREC.
  4. [24]24. Andrew Trask, Felix Hill, Scott Reed, Jack W. Rae, Chris Dyer, and Phil Blunsom. 2018. Neural arith- metic logic units. In NeurIPS.
  5. [29]29. Adams Wei Yu, David Dohan, Minh-Thang Luong, Rui Zhao, Kai Chen, Mohammad Norouzi, and Quoc V. Le. 2018. QANet: Combining local convolution with global self-attention for reading comprehen- sion. In ICLR.

Cited by

Metadata

Authors
Wallace, E., Wang, Y., Li, S., Singh, S., & Gardner, M.
Affiliations
  • Allen Institute for Artificial Intelligence (US)
  • Peking University (CN)
  • University of California, Irvine (US)
Venue
Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)
Published
2019
Pages
5306-5314
OpenAlex
W2974596953
Cited by
227
References
28
Citations by year
2026: 9 · 2025: 16 · 2024: 17 · 2023: 38 · 2022: 41 · 2021: 61 · 2020: 43 · 2019: 2
FWCI
14.06
Type
conference paper
Language
en
Open access
Licence
cc-by
Locations
  • Conference on Empirical Methods in Natural Language Processing (EMNLP) (publishedVersion, cc-by)
  • arXiv (Cornell University) (submittedVersion)
  • arXiv (Cornell University)
  • Greater South Information System
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
  • Physical Sciences › Computer Science › Artificial Intelligence › Natural Language Processing Techniques
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Multimodal Machine Learning Applications
Keywords
numeracy, numerical reasoning, word embeddings, language models

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.