The DOI matches its Crossref record; 61 of the 66 references it cites were found in a registry or an index, and 1 are not checked yet.
A Systematic Assessment of Syntactic Generalization in Neural Language Models
Open accessAbstract
While state-of-the-art neural network models continue to achieve lower perplexity scores on language modeling benchmarks, it remains unknown whether optimizing for broad-coverage predictive performance leads to human-like syntactic knowledge.Furthermore, existing work has not provided a clear picture about the model properties required to produce proper syntactic generalizations.We present a systematic evaluation of the syntactic knowledge of neural language models, testing 20 combinations of model types and data sizes on a set of 34 English-language syntactic test suites.We find substantial differences in syntactic generalization performance by model architecture, with sequential models underperforming other architectures.Factorially manipulating model architecture and training dataset size (1M-40M words), we find that variability in syntactic generalization performance is substantially greater by architecture than by dataset size for the corpora tested in our experiments.Our results also reveal a dissociation between perplexity and syntactic generalization performance.
Overview
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source doesn’t support the claim as written.
In the citing paper
“It is evaluated as the model with outperforming bilingual evaluation understudy (BLEU) score among four state-of-art text generation methods (RNN, GAN, GPT, and CTRL) [66] and the best syntactic generalization score among trained (LSTM, ON-LSTM, Transformer, n-gram) and ofthe-shelf (GPT, JRNN, Transformer-XL) models [46].”
From Synthetic Lies: Understanding AI-Generated Misinformation and Evaluating Algorithmic and Human Solutions (2023) · cites this paper as [46]
Checked by a reasoning model — may be wrong.
Experimental — accuracy not yet measured
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
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017) · arXiv · doi:10.48550/arxiv.1706.03762 · Cited by 120,404
Hochreiter, S., & Schmidhuber, J. (1997) · Neural Computation · doi:10.1162/neco.1997.9.8.1735 · Cited by 102,318
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
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
(2017) · Cited by 11,625
Building a Large Annotated Corpus of English: The Penn Treebank
Marcus, M. (1993) · doi:10.21236/ada273556 · Cited by 7,405
XLNet: Generalized Autoregressive Pretraining for Language Understanding
Yang, Z., Dai, Z., Yang, Y., Carbonell, J., Salakhutdinov, R., & Le, Q. V. (2019) · arXiv · doi:10.48550/arxiv.1906.08237 · Cited by 6,687
GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understanding
Wang, A., Singh, A., Michael, J., Hill, F., Levy, O., & Bowman, S. (2018) · Proceedings of the 2018 EMNLP Workshop BlackboxNLP: Analyzing and Interpreting Neural Networks for NLP · doi:10.18653/v1/w18-5446 · Cited by 4,222
Constraints on variables in syntax
Ross, J. R. (1967) · DSpace@MIT (Massachusetts Institute of Technology) · Cited by 4,102
Loper, E., & Bird, S. (2002) · doi:10.3115/1118108.1118117 · Cited by 3,323
Flagged first, then by citations
4 of the 66 references couldn’t be matched to a known paper.
Research works printed but not linked: 6
- Andrew Carnie. 2012. Syntax: A generative introduc- tion, volume 18. John Wiley & Sons.
- Anastasia Giannakidou. 2011. Negative and positive polarity items: Variation, licensing, and composi- tionality. In Semantics: An international hand- book of natural language meaning, volume 3, pages 1660-1712. Berlin: Mouton de Gruyter.
- Mario Giulianelli, Jack Harding, Florian Mohnert, Dieuwke Hupkes, and Willem Zuidema. 2018. Un- der the hood: Using diagnostic classifiers to in- vestigate and improve how language models track agreement information. In Proceedings of the 2018 EMNLP Workshop BlackboxNLP: Analyzing and In- terpreting Neural Networks for NLP, pages 240- 248.
- Tanya Reinhart. 1981. Definite NP anaphora and c- command domains. Linguistic Inquiry, 12(4):605- 635.
- Ethan Wilcox, Jon Gauthier, Jennifer Hu, Peng Qian, and Roger P. Levy. 2020. Evaluating neural net- works as models of human online language process- ing. In Proceedings of the 42nd Meeting of the Cog- nitive Science Society (CogSci 2020). To appear.
- Ethan Wilcox, Peng Qian, Richard Futrell, Miguel Ballestros, and Roger P. Levy. 2019b. Structural su- pervision improves learning of non-local grammati- cal dependencies. In Proceedings of the 2019 Con- ference of the North American Chapter of the Asso- ciation for Computational Linguistics: Human Lan- guage Technologies, pages 3302-3312, Minneapo- lis, Minnesota.
Cited by
Sorted by citations
Items 101–137 of 137
Portelance, E., Reddy, S., & O’Donnell, T. J. (2025) · Journal of Memory and Language · doi:10.1016/j.jml.2025.104672 · Cited by 2
How is BERT surprised? Layerwise detection of linguistic anomaliesPreprint
Li, B., Zhu, Z., Thomas, G., Xu, Y., & Rudzicz, F. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2105.07452 · Cited by 2
Language Models Fail to Introspect About Their Knowledge of Language
Song, S., Hu, J., & Mahowald, K. (2025) · arXiv · doi:10.48550/arxiv.2503.07513 · Cited by 2
Self-citationIntroduction: Explainability, AI literacy, and language development
Shin, G.-H., & Parde, N. (2024) · Computer Speech & Language · doi:10.1016/j.csl.2024.101766 · Cited by 2
Language mapping functions: Improving softmax estimation and word embedding quality
Rangriz, E., & Pourahmadi, V. (2021) · Concurrency and Computation Practice and Experience · doi:10.1002/cpe.6464 · Cited by 1
Felhi, G., Roux, J. L., & Seddah, D. (2022) · Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT) · doi:10.18653/v1/2022.naacl-main.423 · Cited by 3
Wijnholds, G., & Moortgat, M. (2023) · Conference on Computational Natural Language Learning · doi:10.18653/v1/2023.conll-1.11 · Cited by 4
Takeda, N., Legaspi, R., Nishimura, Y., Ikeda, K., Minamikawa, A., Plötz, T., & Chernova, S. (2024) · Journal of Ambient Intelligence and Smart Environments · doi:10.3233/ais-230429 · Cited by 1
Quantifying Context With and Without Statistical Language Models
Jacobs, C. L. (2022) · Springer eBooks · doi:10.1007/978-3-031-03945-4_17 · Cited by 1
Enhancing the Transformer Decoder with Transition-based Syntax
Choshen, L., & Abend, O. (2022) · Conference on Computational Natural Language Learning · doi:10.18653/v1/2022.conll-1.27 · Cited by 6
Beyond the Bias: Unveiling the Quality of Implicit Causality Prompt Continuations in Language Models
Sieker, J., Bott, O., Solstad, T., & Zarrieß, S. (2023) · International Conference on Natural Language Generation · doi:10.18653/v1/2023.inlg-main.15 · Cited by 4
Acquiring Linguistic Knowledge from Multimodal Input
Amariucai, T., & Warstadt, A. S. (2023) · BabyLM Challenge · doi:10.18653/v1/2023.conll-babylm.11 · Cited by 1
Chobey, A., Smith, O., Wang, A., & Prasad, G. (2023) · BabyLM Challenge · doi:10.18653/v1/2023.conll-babylm.9 · Cited by 1
Perplexity-based molecule ranking and bias estimation of chemical language modelsPreprint
Moret, M., Grisoni, F., Katzberger, P., & Schneider, G. (2021) · doi:10.33774/chemrxiv-2021-zv6f1-v2 · Cited by 1
Advanced Language Understanding with Syntax-Enhanced TransformerPreprint
Rine, W., Patel, R., & Steve, N. (2023) · Preprints.org · doi:10.20944/preprints202312.1673.v1 · Cited by 1
Quantifying Context With and Without Statistical Language Models
Jacobs, C. L. (2021) · Springer eBooks · doi:10.1007/978-3-030-44982-7_17-1
Multilingual Syntax-aware Language Modeling through Dependency Tree Conversion
Kando, S., Noji, H., & Miyao, Y. (2022) · Workshop on Structured Prediction for NLP (SPNLP) · doi:10.18653/v1/2022.spnlp-1.1
Perplexity-based molecule ranking and bias estimation of chemical language modelsPreprint
Moret, M., Grisoni, F., Katzberger, P., & Schneider, G. (2021) · doi:10.33774/chemrxiv-2021-zv6f1 · Cited by 1
Quantum distributional semantics
Correia, A. D. (2022) · doi:10.33540/631
Meaning creation in novel noun-noun compounds: humans and language modelsPreprint
Chen, P., Poeppel, D., & Zuanazzi, A. (2023) · doi:10.31234/osf.io/gf2r9
Syntax and Geometry of Information
Bailly, R., Leblond, L., & Gábor, K. (2023) · Proceedings of the conference - Association for Computational Linguistics. Meeting · doi:10.18653/v1/2023.acl-long.590
Composition, Attention, or Both?
Yoshida, R., & Oseki, Y. (2022) · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP) · doi:10.18653/v1/2022.findings-emnlp.428
Prange, J., & Chersoni, E. (2023) · Joint Conference on Lexical and Computational Semantics · doi:10.18653/v1/2023.starsem-1.40
How Much Syntactic Supervision is “Good Enough”?
Noji, H., & Oseki, Y. (2023) · Findings of the Association for Computational Linguistics: EACL · doi:10.18653/v1/2023.findings-eacl.173
Beinborn, L., & Hollenstein, N. (2023) · Synthesis lectures on human language technologies · doi:10.1007/978-3-031-43260-6_4
NPIs Aren’t Exactly Easy: Variation in Licensing across Large Language Models
DeCarlo, D., Palmer, W., Wilson, M., & Frank, B. (2023) · BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP (BlackboxNLP) · doi:10.18653/v1/2023.blackboxnlp-1.25
Pushdown Layers: Encoding Recursive Structure in Transformer Language Models
Murty, S., Sharma, P., Andreas, J., & Manning, C. D. (2023) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2023.emnlp-main.195 · Cited by 8
Clark, T. H., Poliak, M., Regev, T. I., Haskins, A. J., Robertson, C. E., & Gibson, E. A. F. (2025) · Cognitive Science · doi:10.1111/cogs.70134
Inducing Grammatical Knowledge from Indirect Evidence in Language Models
Oba, M., Oseki, Y., Fukatsu, A., Haga, A., Ouchi, H., Watanabe, T., & Sugawara, S. (2026) · Journal of Natural Language Processing · doi:10.5715/jnlp.33.109
Study of the abstraction capabilities of neural language models
Li, B. (2023) · theses.fr (ABES) · doi:10.70675/0236e084zee1ez4013z8f5bz6eebee3f0e4e
Dunagan, D. G., Low, D. S., Yue, S., Meyer, L., & Hale, J. T. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.04.20.719609
A Systematic Assessment of Language Models with Linguistic Minimal Pairs in Chinese
Liu, Y., Shen, Y., Zhu, H., Xu, L., Qian, Z., Song, S., Zhang, K., Tang, J., Zhang, P., Yang, B., et al. (2026) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl.a.648 · Cited by 24
Child-Directed Speech Facilitates Semantic Role Learning: A Machine Learning Approach
Huber, E., Bickel, B., & Stoll, S. (2026) · Open Mind · doi:10.1162/opmi.a.351
Catasso, N. (2026) · AI-Linguistica Linguistic Studies on AI-Generated Texts and Discourses · doi:10.62408/ai-ling.v3i1.88 · Cited by 1
Focus shifts in contextual and lexical cue interactions in GPT models
Chung, W., & Koo, K. (2026) · PLoS ONE · doi:10.1371/journal.pone.0351924
Timkey, W., Huang, K.-J., Oh, B.-D., Prasad, G., Arehalli, S., Linzen, T., & Dillon, B. (2026) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2532230123 · Cited by 4
Moreira, R., LaForce, S., Teles, A. S., Harris, F., & Moon, N. W. (2026) · Universal Access in the Information Society · doi:10.1007/s10209-026-01380-1
Metadata
- Authors
- Hu, J., Gauthier, J., Qian, P., Wilcox, E., & Levy, R.
- Affiliations
- Massachusetts Institute of Technology (US)
- Harvard University (US)
- Venue
- Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics
- Published
- 2020
- Pages
- 1725-1744
- ISSN
- 0736-587X (ISSN-L)
- OpenAlex
- W3034510440
- Cited by
- 134
- References
- 66
- Citations by year
- 2026: 11 · 2025: 10 · 2024: 17 · 2023: 40 · 2022: 26 · 2021: 25 · 2020: 5
- FWCI
- 9.17
- Type
- conference paper
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- Locations
- Proceedings of the conference - Association for Computational Linguistics. Meeting (publishedVersion, cc-by)
- Funders
- National Science Foundation
- Open Philanthropy Project
- National Institutes of Health
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Natural Language Processing Techniques
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Social Sciences › Arts and Humanities › Language and Linguistics › Syntax, Semantics, Linguistic Variation
- Keywords
- syntactic generalization, neural language models, syntactic knowledge, model architecture, training data size, perplexity, English syntax, sequential model, language model evaluation
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
