pith. sign in

arxiv: 2606.02953 · v1 · pith:4S6K6R2Xnew · submitted 2026-06-01 · 💻 cs.CL

Linguistic Productivity in Large Language Models: Models Coerce, but do not Preempt

Claire Bonial , Claire Benet Post , Laura Michaelis , Harish Tayyar Madabushi This is my paper

Pith reviewed 2026-06-28 14:11 UTC · model grok-4.3

classification 💻 cs.CL
keywords linguistic productivityentrenchmentpreemptionconstructional coercionnonce wordslarge language modelsusage-based theoriesovergeneralization
0
0 comments X

The pith

Large language models capture entrenchment through coercion with nonce words but show no preemption from absent patterns.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Usage-based theories hold that language productivity is supported by frequent exposure to structures and limited by their consistent absence where expected. The paper tests whether these same frequency signals shape how LLMs generate and interpret language. Experiments show larger models can extend constructions to made-up words when context forces an atypical meaning, reproducing the entrenchment side of the theory. The same models nevertheless produce overgeneralizations of semantically acceptable patterns that never occurred in their training data, indicating they do not register or apply negative evidence in the way preemption requires.

Core claim

Across model sizes and architectures, LLMs reproduce constructional productivity via entrenchment when a broader frame coerces an atypical reading of a nonce word, yet they continue to overgeneralize patterns that are semantically acceptable but unattested, showing that statistical preemption does not constrain their output.

What carries the argument

The contrast between entrenchment, driven by high-frequency usage of a construction, and preemption, driven by consistent non-occurrence in contexts where the construction might otherwise appear, tested through nonce-word substitution in coercion frames.

If this is right

  • Larger models increasingly exhibit entrenchment effects that allow coerced interpretations with novel lexical items.
  • Models of any size fail to block overgeneralization of unattested but semantically coherent patterns.
  • Statistical absence alone does not function as a learning signal for LLMs in the manner predicted by preemption accounts.
  • The dissociation between coercion success and preemption failure holds across different model architectures.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • This pattern suggests LLMs may need explicit mechanisms for registering negative evidence if they are to match human-like avoidance of certain generalizations.
  • The result points to a possible test: whether targeted exposure to unattested constructions paired with corrective signals reduces overgeneralization in subsequent generations.
  • It raises the question of whether other statistical or architectural features, beyond raw frequency counts, could supply the missing preemption effect.

Load-bearing premise

The specific nonce-word tasks and construction frames used here validly isolate the same entrenchment and preemption mechanisms that usage-based theories attribute to human speakers.

What would settle it

A controlled test in which models trained or prompted with explicit negative evidence for a semantically acceptable but unattested construction subsequently stop producing that construction at rates significantly above baseline.

read the original abstract

Usage-based theories of grammars posit that creative productivity of the structures of language is both bolstered and constrained by two distinct frequency signals: entrenchment, stemming from high frequency usage, and preemption, stemming from having never observed a particular linguistic structure in a context where one might expect that structure to appear. Large Language Models are also usage-based, in the sense that the structures of language are learned through exposure to vast amounts of text. Here, we test whether or not the opposing statistical forces of entrenchment and preemption also encourage and constrain linguistic productivity in LLMs. We demonstrate across model architectures that larger models recognize and can reproduce with nonce words constructional productivity (entrenchment) in cases of coercion, wherein the broader constructional context coerces an atypical interpretation of a lexical item. However, we also show that even the largest models do not extend negative evidence to novel language, and statistical preemption does not enable models to avoid overgeneralization of patterns that are semantically felicitous, but never observed in data.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 0 minor

Summary. The paper claims that LLMs exhibit entrenchment-driven constructional productivity (via coercion with nonce words) that scales with model size, but lack preemption effects from negative evidence, failing to block overgeneralization on semantically felicitous but unattested patterns; this dissociation is presented as holding across architectures and as evidence that statistical preemption does not constrain LLM productivity in the manner predicted by usage-based theories.

Significance. If the dissociation is robustly demonstrated, the result would bear on whether LLMs implement the two distinct frequency signals posited in usage-based grammar, with potential implications for cognitive modeling of productivity. The nonce-word design is a standard tool for testing generalization and is a positive feature when properly controlled.

major comments (2)
  1. [Abstract] Abstract: results are asserted across architectures with no accompanying details on test constructions, statistical controls, sample sizes, or the operationalization of overgeneralization; without these elements the central empirical claim cannot be evaluated.
  2. [Experimental tasks] The reported failure to avoid overgeneralization on unattested but felicitous patterns is taken to demonstrate absence of preemption; however, this inference requires showing that the nonce-word frames and prompting regime provide sufficient negative evidence and isolate preemption from architecture-specific limits on representing absence, which is not addressed.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their detailed and constructive comments. We address each major comment point by point below, indicating planned revisions where appropriate.

read point-by-point responses

  1. Referee: [Abstract] Abstract: results are asserted across architectures with no accompanying details on test constructions, statistical controls, sample sizes, or the operationalization of overgeneralization; without these elements the central empirical claim cannot be evaluated.

    Authors: We agree that the abstract is highly condensed. The full manuscript specifies the constructions (coercion frames with nonce words), controls (model size and architecture comparisons), sample sizes (multiple LLMs and prompt variants), and operationalization (preference rates for attested vs. unattested patterns). We will revise the abstract to include a concise reference to these elements. revision: partial

  2. Referee: [Experimental tasks] The reported failure to avoid overgeneralization on unattested but felicitous patterns is taken to demonstrate absence of preemption; however, this inference requires showing that the nonce-word frames and prompting regime provide sufficient negative evidence and isolate preemption from architecture-specific limits on representing absence, which is not addressed.

    Authors: The nonce-word coercion design follows established usage-based methods to supply contexts where preemption from negative evidence would be expected if utilized. Testing across architectures and sizes helps separate general statistical effects from model-specific constraints. We will add explicit discussion in the Methods and Discussion sections on the prompting regime's provision of negative evidence and note limitations in fully isolating preemption from representational factors. revision: yes

Circularity Check

0 steps flagged

Empirical evaluation with no derivation chain or self-referential reductions

full rationale

The paper reports experimental results from prompting LLMs with nonce words in specific constructional frames to test entrenchment (via coercion) versus preemption effects. No equations, parameters, or derivations appear in the abstract or described content; outcomes are direct model generations compared to linguistic expectations from usage-based theories. No self-citations function as load-bearing uniqueness theorems, no fitted inputs are relabeled as predictions, and no ansatzes or renamings reduce claims to inputs by construction. The central dissociation between coercion recognition and failure to use negative evidence follows from observed outputs against external benchmarks, rendering the work self-contained.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The paper tests an existing linguistic theory on LLMs rather than deriving new constants or introducing new entities; background assumptions are standard in usage-based linguistics.

axioms (1)
  • domain assumption Usage-based theories of grammars posit that creative productivity of the structures of language is both bolstered and constrained by two distinct frequency signals: entrenchment and preemption.
    This is the core theoretical premise the experiments are designed to test in LLMs.

pith-pipeline@v0.9.1-grok · 5721 in / 1312 out tokens · 29735 ms · 2026-06-28T14:11:06.475662+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

297 extracted references · 54 canonical work pages · 3 internal anchors

  1. [1]

    , editor=

    Qi, Peng and Zhang, Yuhao and Zhang, Yuhui and Bolton, Jason and Manning, Christopher D. , editor=. Stanza:. Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics: System Demonstrations , publisher=. 2020 , month=jul, pages=. doi:10.18653/v1/2020.acl-demos.14 , abstractNote=

    work page doi:10.18653/v1/2020.acl-demos.14 2020

  2. [2]

    Devlin, Jacob and Chang, Ming-Wei and Lee, Kenton and Toutanova, Kristina , editor=. B. 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) , publisher=. 2019 , month=jun, pages=. doi:10.18653/v1/N19-1423 , abstractNote=

    work page doi:10.18653/v1/n19-1423 2019

  3. [3]

    Mahowald, Kyle , year=. A. Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics , publisher=

  4. [4]

    “Construction after

    Jackendoff, Ray , year=. “Construction after. Language , publisher=

  5. [5]

    The better your

    Weissweiler, Leonie and Hofmann, Valentin and Köksal, Abdullatif and Schütze, Hinrich , editor=. The better your. Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing , publisher=. 2022 , month=dec, pages=. doi:10.18653/v1/2022.emnlp-main.746 , abstractNote=

    work page doi:10.18653/v1/2022.emnlp-main.746 2022

  6. [6]

    Findings of the B aby LM Challenge: Sample-Efficient Pretraining on Developmentally Plausible Corpora

    Warstadt, Alex and Mueller, Aaron and Choshen, Leshem and Wilcox, Ethan and Zhuang, Chengxu and Ciro, Juan and Mosquera, Rafael and Paranjabe, Bhargavi and Williams, Adina and Linzen, Tal and Cotterell, Ryan. Findings of the B aby LM Challenge: Sample-Efficient Pretraining on Developmentally Plausible Corpora. Proceedings of the BabyLM Challenge at the 27...

    work page doi:10.18653/v1/2023.conll-babylm.1 2023

  7. [7]

    Constructions are Revealed in Word Distributions

    Rozner, Joshua and Weissweiler, Leonie and Mahowald, Kyle and Shain, Cory. Constructions are Revealed in Word Distributions. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing. 2025

    2025

  8. [8]

    B aby LM ' s First Constructions: Causal interventions provide a signal of learning

    Rozner, Joshua and Weissweiler, Leonie and Shain, Cory. B aby LM ' s First Constructions: Causal interventions provide a signal of learning. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing. 2025

    2025

  9. [9]

    arXiv preprint arXiv:1907.11692 , year=

    Roberta: A robustly optimized bert pretraining approach , author=. arXiv preprint arXiv:1907.11692 , year=

    Pith/arXiv arXiv 1907

  10. [10]

    Evaluating C x G Generalisation in LLM s via Construction-Based NLI Fine Tuning

    Mackintosh, Tom and Tayyar Madabushi, Harish and Bonial, Claire. Evaluating C x G Generalisation in LLM s via Construction-Based NLI Fine Tuning. Proceedings of the Second International Workshop on Construction Grammars and NLP. 2025

    2025

  11. [11]

    UC xn: Typologically Informed Annotation of Constructions Atop U niversal D ependencies

    Weissweiler, Leonie and B. UC xn: Typologically Informed Annotation of Constructions Atop U niversal D ependencies. Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024). 2024

    2024

  12. [12]

    Construction Identification and Disambiguation Using BERT : A Case Study of NPN

    Scivetti, Wesley and Schneider, Nathan. Construction Identification and Disambiguation Using BERT : A Case Study of NPN. Proceedings of the 29th Conference on Computational Natural Language Learning. 2025. doi:10.18653/v1/2025.conll-1.24

    work page doi:10.18653/v1/2025.conll-1.24 2025

  13. [13]

    Do Construction Distributions Shape Formal Language Learning In G erman B aby LM s?

    Bunzeck, Bastian and Duran, Daniel and Zarrie , Sina. Do Construction Distributions Shape Formal Language Learning In G erman B aby LM s?. Proceedings of the 29th Conference on Computational Natural Language Learning. 2025. doi:10.18653/v1/2025.conll-1.12

    work page doi:10.18653/v1/2025.conll-1.12 2025

  14. [14]

    and Varoquaux, G

    Pedregosa, F. and Varoquaux, G. and Gramfort, A. and Michel, V. and Thirion, B. and Grisel, O. and Blondel, M. and Prettenhofer, P. and Weiss, R. and Dubourg, V. and Vanderplas, J. and Passos, A. and Cournapeau, D. and Brucher, M. and Perrot, M. and Duchesnay, E. , journal=. Scikit-learn: Machine Learning in

  15. [15]

    Unpacking Let Alone: Human-Scale Models Generalize to a Rare Construction in Form but not Meaning

    Scivetti, Wesley and Aoyama, Tatsuya and Wilcox, Ethan and Schneider, Nathan. Unpacking Let Alone: Human-Scale Models Generalize to a Rare Construction in Form but not Meaning. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing. 2025

    2025

  16. [16]

    Language Models Learn Rare Phenomena from Less Rare Phenomena:

    Misra, Kanishka and Mahowald, Kyle , editor =. Language Models Learn Rare Phenomena from Less Rare Phenomena:. Proceedings of the 2024. 2024 , pages =. doi:10.18653/v1/2024.emnlp-main.53 , abstract =

    work page doi:10.18653/v1/2024.emnlp-main.53 2024

  17. [17]

    2024 , eprint=

    Time Travel in LLMs: Tracing Data Contamination in Large Language Models , author=. 2024 , eprint=

    2024

  18. [18]

    Modeling Semantic Containment and Exclusion in Natural Language Inference

    MacCartney, Bill and Manning, Christopher D. Modeling Semantic Containment and Exclusion in Natural Language Inference. Proceedings of the 22nd International Conference on Computational Linguistics (Coling 2008). 2008

    2008

  19. [19]

    Tayyar Madabushi, Harish and Romain, Laurence and Divjak, Dagmar and Milin, Petar , editor=. Cx. Proceedings of the 28th International Conference on Computational Linguistics , publisher=. 2020 , month=dec, pages=. doi:10.18653/v1/2020.coling-main.355 , abstractNote=

    work page doi:10.18653/v1/2020.coling-main.355 2020

  20. [20]

    Belinkov

    Probing. Computational Linguistics , author=. 2022 , month=apr, pages=. doi:10.1162/coli_a_00422 , abstractNote=

    work page internal anchor Pith review doi:10.1162/coli_a_00422 2022

  21. [21]

    BL i MP : The Benchmark of Linguistic Minimal Pairs for E nglish

    Transactions of the Association for Computational Linguistics , author=. 2020 , month=jul, pages=. doi:10.1162/tacl_a_00321 , abstractNote=

    work page doi:10.1162/tacl_a_00321 2020

  22. [22]

    Yaghoobzadeh, Yadollah and Kann, Katharina and Hazen, T. J. and Agirre, Eneko and Schütze, Hinrich , editor=. Probing for. Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics , publisher=. 2019 , month=jul, pages=. doi:10.18653/v1/P19-1574 , abstractNote=

    work page doi:10.18653/v1/p19-1574 2019

  23. [23]

    Vulić, Ivan and Ponti, Edoardo Maria and Litschko, Robert and Glavaš, Goran and Korhonen, Anna , editor=. Probing. Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP) , publisher=. 2020 , month=nov, pages=. doi:10.18653/v1/2020.emnlp-main.586 , abstractNote=

    work page doi:10.18653/v1/2020.emnlp-main.586 2020

  24. [25]

    Zhou, Yichu and Srikumar, Vivek , editor=. Direct. Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies , publisher=. 2021 , month=jun, pages=. doi:10.18653/v1/2021.naacl-main.401 , abstractNote=

    work page doi:10.18653/v1/2021.naacl-main.401 2021

  25. [26]

    Aoyama, Tatsuya and Schneider, Nathan , editor=. Probe-. Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies: Student Research Workshop , publisher=. 2022 , month=jul, pages=. doi:10.18653/v1/2022.naacl-srw.25 , abstractNote=

    work page doi:10.18653/v1/2022.naacl-srw.25 2022

  26. [27]

    Hoffmann, J., Borgeaud, S., Mensch, A., Buchatskaya, E., Cai, T., Rutherford, E., Casas, D

    Hewitt, John and Liang, Percy , editor=. Designing and. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP) , publisher=. 2019 , month=nov, pages=. doi:10.18653/v1/D19-1275 , abstractNote=

    work page doi:10.18653/v1/d19-1275 2019

  27. [28]

    Jawahar, Ganesh and Sagot, Benoît and Seddah, Djamé , editor=. What. Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics , publisher=. 2019 , month=jul, pages=. doi:10.18653/v1/P19-1356 , abstractNote=

    work page doi:10.18653/v1/p19-1356 2019

  28. [29]

    Counterfactual

    Ravfogel, Shauli and Prasad, Grusha and Linzen, Tal and Goldberg, Yoav , editor=. Counterfactual. Proceedings of the 25th Conference on Computational Natural Language Learning , publisher=. 2021 , month=nov, pages=. doi:10.18653/v1/2021.conll-1.15 , abstractNote=

    work page doi:10.18653/v1/2021.conll-1.15 2021

  29. [30]

    What Does BERT Look at? An Analysis of BERT ' s Attention

    Clark, Kevin and Khandelwal, Urvashi and Levy, Omer and Manning, Christopher D. , editor=. What. Proceedings of the 2019 ACL Workshop BlackboxNLP: Analyzing and Interpreting Neural Networks for NLP , publisher=. 2019 , month=aug, pages=. doi:10.18653/v1/W19-4828 , abstractNote=

    work page doi:10.18653/v1/w19-4828 2019

  30. [31]

    Karidi, Taelin and Zhou, Yichu and Schneider, Nathan and Abend, Omri and Srikumar, Vivek , editor=. Putting. Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing , publisher=. 2021 , month=nov, pages=. doi:10.18653/v1/2021.emnlp-main.806 , abstractNote=

    work page doi:10.18653/v1/2021.emnlp-main.806 2021

  31. [32]

    Liu, Matt Gardner, Yonatan Belinkov, Matthew E

    Liu, Nelson F. and Gardner, Matt and Belinkov, Yonatan and Peters, Matthew E. and Smith, Noah A. , editor=. Linguistic. 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) , publisher=. 2019 , month=jun, pages=. doi:10.18653/v1/N19-...

    work page doi:10.18653/v1/n19-1112 2019

  32. [33]

    Probing for

    Conia, Simone and Navigli, Roberto , editor=. Probing for. Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) , publisher=. 2022 , month=may, pages=. doi:10.18653/v1/2022.acl-long.316 , abstractNote=

    work page doi:10.18653/v1/2022.acl-long.316 2022

  33. [34]

    Probing for semantic evidence of composition by means of simple classification tasks , url=

    Ettinger, Allyson and Elgohary, Ahmed and Resnik, Philip , year=. Probing for semantic evidence of composition by means of simple classification tasks , url=. doi:10.18653/v1/W16-2524 , booktitle=

    work page doi:10.18653/v1/w16-2524

  34. [35]

    Amnesic probing: Behavioral explanation with amnesic counterfactuals.Transactions of the Association for Computational Linguistics, 9:160–175, 2021

    Amnesic. Transactions of the Association for Computational Linguistics , author=. 2021 , month=mar, pages=. doi:10.1162/tacl_a_00359 , abstractNote=

    work page doi:10.1162/tacl_a_00359 2021

  35. [36]

    and Pimentel, Tiago and Saphra, Naomi and Cotterell, Ryan , editor=

    White, Jennifer C. and Pimentel, Tiago and Saphra, Naomi and Cotterell, Ryan , editor=. A. Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies , publisher=. 2021 , month=jun, pages=. doi:10.18653/v1/2021.naacl-main.12 , abstractNote=

    work page doi:10.18653/v1/2021.naacl-main.12 2021

  36. [37]

    Doran, & T

    Hewitt, John and Manning, Christopher D. , editor=. A. 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) , publisher=. 2019 , month=jun, pages=. doi:10.18653/v1/N19-1419 , abstractNote=

    work page doi:10.18653/v1/n19-1419 2019

  37. [38]

    Tenney, Ian and Das, Dipanjan and Pavlick, Ellie , editor=. B. Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics , publisher=. 2019 , month=jul, pages=. doi:10.18653/v1/P19-1452 , abstractNote=

    work page doi:10.18653/v1/p19-1452 2019

  38. [39]

    Goldberg, Adele E. , year=. Constructions:

  39. [40]

    Croft, William , year=. Radical

  40. [41]

    Goldberg, Adele E. , year=. Constructions at

  41. [42]

    Tseng, Yu-Hsiang and Shih, Cing-Fang and Chen, Pin-Er and Chou, Hsin-Yu and Ku, Mao-Chang and Hsieh, Shu-Kai , editor=. Cx. Proceedings of the Thirteenth Language Resources and Evaluation Conference , publisher=. 2022 , month=jun, pages=

    2022

  42. [44]

    Veenboer, Tim and Bloem, Jelke , editor=. Using. Findings of the Association for Computational Linguistics: ACL 2023 , publisher=. 2023 , month=jul, pages=. doi:10.18653/v1/2023.findings-acl.819 , abstractNote=

    work page doi:10.18653/v1/2023.findings-acl.819 2023

  43. [45]

    Chronis, Gabriella and Mahowald, Kyle and Erk, Katrin , editor=. A. Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) , publisher=. 2023 , month=jul, pages=. doi:10.18653/v1/2023.acl-long.14 , abstractNote=

    work page doi:10.18653/v1/2023.acl-long.14 2023

  44. [46]

    Pannitto, Ludovica and Herbelot, Aurélie , editor=. C. Proceedings of the First International Workshop on Construction Grammars and NLP (CxGs+NLP, GURT/SyntaxFest 2023) , publisher=. 2023 , month=mar, pages=

    2023

  45. [47]

    Assessing BERT's Syntactic Abilities

    Goldberg, Yoav , year=. Assessing. doi:10.48550/arXiv.1901.05287 , abstractNote=

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.1901.05287 1901

  46. [48]

    Reconstruction

    Kim, Najoung and Khilnani, Jatin and Warstadt, Alex and Qaddoumi, Abed , year=. Reconstruction. doi:10.48550/arXiv.2212.10792 , abstractNote=

    work page doi:10.48550/arxiv.2212.10792

  47. [49]

    Kulmizev, Artur and Ravishankar, Vinit and Abdou, Mostafa and Nivre, Joakim , editor=. Do. Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics , publisher=. 2020 , month=jul, pages=. doi:10.18653/v1/2020.acl-main.375 , abstractNote=

    work page doi:10.18653/v1/2020.acl-main.375 2020

  48. [50]

    Wolf, Thomas and Debut, Lysandre and Sanh, Victor and Chaumond, Julien and Delangue, Clement and Moi, Anthony and Cistac, Pierric and Rault, Tim and Louf, Rémi and Funtowicz, Morgan and Davison, Joe and Shleifer, Sam and von Platen, Patrick and Ma, Clara and Jernite, Yacine and Plu, Julien and Xu, Canwen and Scao, Teven Le and Gugger, Sylvain and Drame, M...

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.1910.03771 1910

  49. [51]

    Lin, Yongjie and Tan, Yi Chern and Frank, Robert , editor=. Open. Proceedings of the 2019 ACL Workshop BlackboxNLP: Analyzing and Interpreting Neural Networks for NLP , publisher=. 2019 , month=aug, pages=. doi:10.18653/v1/W19-4825 , abstractNote=

    work page doi:10.18653/v1/w19-4825 2019

  50. [52]

    Information-

    Pimentel, Tiago and Valvoda, Josef and Maudslay, Rowan Hall and Zmigrod, Ran and Williams, Adina and Cotterell, Ryan , editor=. Information-. Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics , publisher=. 2020 , month=jul, pages=. doi:10.18653/v1/2020.acl-main.420 , abstractNote=

    work page doi:10.18653/v1/2020.acl-main.420 2020

  51. [53]

    Goldberg, Adele E. , year=. Usage-based constructionist approaches and Large Language Models , url=. doi:10.31234/osf.io/8bmwz , abstractNote=

    work page doi:10.31234/osf.io/8bmwz

  52. [54]

    Constructing a Language: A Usage-Based Theory of Language Acquisition , ISBN=

    Tomasello, Michael , year=. Constructing a Language: A Usage-Based Theory of Language Acquisition , ISBN=

  53. [55]

    Lingbuzz , author=

    Modern language models refute Chomsky’s approach to language , volume=. Lingbuzz , author=. 2023 , language=

    2023

  54. [56]

    Mortensen, David and Levin, Lori and Schütze, Hinrich , editor=

    Weissweiler, Leonie and He, Taiqi and Otani, Naoki and R. Mortensen, David and Levin, Lori and Schütze, Hinrich , editor=. Construction Grammar Provides Unique Insight into Neural Language Models , url=. Proceedings of the First International Workshop on Construction Grammars and NLP (CxGs+NLP, GURT/SyntaxFest 2023) , publisher=. 2023 , month=mar, pages=

    2023

  55. [57]

    and Levin, Lori , editor=

    Zhou, Shijia and Weissweiler, Leonie and He, Taiqi and Schütze, Hinrich and Mortensen, David R. and Levin, Lori , editor=. Constructions. Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024) , publisher=. 2024 , month=may, pages=

    2024

  56. [58]

    and Izrailevitch, Valentina and Xiao, Yunze and Schütze, Hinrich and Weissweiler, Leonie , editor=

    Mortensen, David R. and Izrailevitch, Valentina and Xiao, Yunze and Schütze, Hinrich and Weissweiler, Leonie , editor=. Verbing Weirds Language (Models): Evaluation of English Zero-Derivation in Five LLMs , url=. Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024) , publ...

    2024

  57. [59]

    A C onstruction G rammar C orpus of V arying S chematicity: A D ataset for the E valuation of A bstractions in L anguage M odels

    Bonial, Claire and Tayyar Madabushi, Harish. A C onstruction G rammar C orpus of V arying S chematicity: A D ataset for the E valuation of A bstractions in L anguage M odels. Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024). 2024

    2024

  58. [60]

    Advances in Neural Information Processing Systems , author=

    Chain-of-. Advances in Neural Information Processing Systems , author=. 2022 , month=dec, pages=

    2022

  59. [61]

    Building

    Schäfer, Roland and Bildhauer, Felix , editor=. Building. Proceedings of the Eighth International Conference on Language Resources and Evaluation (LREC’12) , publisher=. 2012 , month=may, pages=

    2012

  60. [62]

    Proceedings of the 3rd Workshop on Challenges in the Management of Large Corpora , author=

    Processing and querying large web corpora with the. Proceedings of the 3rd Workshop on Challenges in the Management of Large Corpora , author=. 2015 , pages=

    2015

  61. [63]

    2016 , eprint=

    Neural Machine Translation by Jointly Learning to Align and Translate , author=. 2016 , eprint=

    2016

  62. [64]

    A Primer in BERT ology: What We Know About How BERT Works

    Rogers, Anna and Kovaleva, Olga and Rumshisky, Anna. A Primer in BERT ology: What We Know About How BERT Works. Transactions of the Association for Computational Linguistics. 2020. doi:10.1162/tacl_a_00349

    work page doi:10.1162/tacl_a_00349 2020

  63. [65]

    Proceedings of the National Academy of Sciences , volume=

    Emergent linguistic structure in artificial neural networks trained by self-supervision , author=. Proceedings of the National Academy of Sciences , volume=. 2020 , publisher=

    2020

  64. [66]

    Why C an GPT L earn I n- C ontext? L anguage M odels S ecretly P erform G radient Descent as M eta- O ptimizers

    Dai, Damai and Sun, Yutao and Dong, Li and Hao, Yaru and Ma, Shuming and Sui, Zhifang and Wei, Furu. Why C an GPT L earn I n- C ontext? L anguage M odels S ecretly P erform G radient Descent as M eta- O ptimizers. Findings of the Association for Computational Linguistics: ACL 2023. 2023. doi:10.18653/v1/2023.findings-acl.247

    work page doi:10.18653/v1/2023.findings-acl.247 2023

  65. [67]

    Dai and Quoc V Le , booktitle=

    Jason Wei and Maarten Bosma and Vincent Zhao and Kelvin Guu and Adams Wei Yu and Brian Lester and Nan Du and Andrew M. Dai and Quoc V Le , booktitle=. Finetuned. 2022 , url=

    2022

  66. [68]

    Sheng Lu and Irina Bigoulaeva and Rachneet Sachdeva and Harish Tayyar Madabushi and Iryna Gurevych , year=. Are. 2309.01809 , archivePrefix=

    arXiv

  67. [69]

    Sainz, Oscar and Campos, Jon Ander and García-Ferrero, Iker and Etxaniz, Julen and Agirre, Eneko , year=. Did

  68. [70]

    NLP E valuation in trouble: O n the N eed to M easure LLM D ata C ontamination for each B enchmark

    Sainz, Oscar and Campos, Jon and Garc \' a-Ferrero, Iker and Etxaniz, Julen and de Lacalle, Oier Lopez and Agirre, Eneko. NLP E valuation in trouble: O n the N eed to M easure LLM D ata C ontamination for each B enchmark. Findings of the Association for Computational Linguistics: EMNLP 2023. 2023. doi:10.18653/v1/2023.findings-emnlp.722

    work page doi:10.18653/v1/2023.findings-emnlp.722 2023

  69. [71]

    Lewis, Martha and Mitchell, Melanie , journal=. Using

  70. [72]

    Data Contamination: From Memorization to Exploitation

    Magar, Inbal and Schwartz, Roy. Data Contamination: From Memorization to Exploitation. Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers). 2022. doi:10.18653/v1/2022.acl-short.18

    work page doi:10.18653/v1/2022.acl-short.18 2022

  71. [73]

    Advances in neural information processing systems , volume=

    Language models are few-shot learners , author=. Advances in neural information processing systems , volume=

  72. [74]

    Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing , year=

    A large annotated corpus for learning natural language inference , author=. Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing , year=

    2015

  73. [75]

    A brief history of natural logic , author=

  74. [76]

    SemEval-2014 Task 1: Evaluation of Compositional Distributional Semantic Models on Full Sentences through Semantic Relatedness and Textual Entailment , doi =

    Marelli, Marco and Bentivogli, Luisa and Baroni, Marco and Bernardi, Raffaella and Menini, Stefano and Zamparelli, Roberto , year =. SemEval-2014 Task 1: Evaluation of Compositional Distributional Semantic Models on Full Sentences through Semantic Relatedness and Textual Entailment , doi =

    2014

  75. [77]

    Proceedings of the 18th International Workshop on Semantic Evaluation (SemEval-2024) , pages=

    Nycu-nlp at semeval-2024 task 2: Aggregating large language models in biomedical natural language inference for clinical trials , author=. Proceedings of the 18th International Workshop on Semantic Evaluation (SemEval-2024) , pages=

    2024

  76. [78]

    The Cambridge handbook of child language , pages=

    The usage-based theory of language acquisition , author=. The Cambridge handbook of child language , pages=. 2009 , publisher=

    2009

  77. [79]

    Chomsky, Noam , year=. The

  78. [80]

    Computational learning of construction grammars , volume=

    Dunn, Jonathan , year=. Computational learning of construction grammars , volume=. Language and Cognition , publisher=. doi:10.1017/langcog.2016.7 , abstractNote=

    work page doi:10.1017/langcog.2016.7 2016

  79. [81]

    Exploring the Constructicon: Linguistic Analysis of a Computational CxG , url=

    Dunn, Jonathan , editor=. Exploring the Constructicon: Linguistic Analysis of a Computational CxG , url=. Proceedings of the First International Workshop on Construction Grammars and NLP (CxGs+NLP, GURT/SyntaxFest 2023) , publisher=. 2023 , month=mar, pages=

    2023

  80. [82]

    Language and Cognitive Processes , volume=

    Evidence for automatic accessing of constructional meaning: Jabberwocky sentences prime associated verbs , author=. Language and Cognitive Processes , volume=. 2013 , publisher=

    2013

Showing first 80 references.