Pith. sign in

REVIEW 4 major objections 4 minor 1 cited by

Explain-then-Process: Using Grammar Prompting to Enhance Grammatical Acceptability Judgments

T0 review · 4 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Grammar prompting — feeding a model's own explanation of a syntactic rule back to it before a minimal-pair judgment — closes most of the accuracy gap between small and large language models on English, Chinese, and Russian grammaticality…

desk verdict An honest, controlled study of feeding LLM-written grammar rules back to SLMs, whose headline 56% gap reduction is a subset artifact but whose method and controls deserve a careful referee. read the letter →

arxiv 2506.02302 v1 pith:MHJOFBXD submitted 2025-06-02 cs.CL cs.AI

Russell Scheinberg , Ameeta Agrawal , Amber Shore , So Young Lee This is my paper · ORCID
classification cs.CLcs.AI
keywords grammarpromptingexplain-then-processgrammaticalacceptabilityjudgmentsminimalpairschain-of-thoughtreasoningsmalllanguagemodelsmultilingualevaluationsyntactic
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper proposes "grammar prompting," a two-step recipe: first a large language model writes a concise explanation of a syntactic phenomenon, then that explanation is inserted into the prompt of the model that must decide which sentence of a minimal pair is grammatical. The authors' central claim is that this "explain-then-process" loop turns latent rule knowledge into applied rule use, and that it helps smaller models most. On selected challenging paradigms from BLiMP (English), SLING (Chinese), and RuBLiMP (Russian), grammar prompting alone cuts the average LLM-SLM accuracy gap from 13.0 to 10.4 percentage points, and grammar prompting combined with chain-of-thought reasoning cuts it to 5.8 points, a 56% reduction. The authors argue this lets low-cost smaller models approach frontier-LLM performance on structured linguistic judgments without any training or task-specific tuning. The headline gap numbers are computed on manually chosen hard paradigm subsets, not on full benchmarks.

What carries the argument

The load-bearing mechanism is the "grammar prompt": a few-hundred-word, model-generated explanation of the target syntactic phenomenon, written for a novice learner, with full example sentences deliberately excluded. It is combined with the minimal-pair judgment prompt, optionally with chain-of-thought reasoning (GP+CoT). The explanation supplies the linguistic categories and constraints (for example, that "only" licenses the negative-polarity item "ever" only when it scopes over the whole phrase), redirecting the model's reasoning from semantic paraphrase — the failure mode the paper diagnoses — to structural rule application. Beginner-oriented explanations outperform expert-oriented ones, and a single relevant explanation beats both an irrelevant control explanation and a multi-explanation "textbook" condition.

What would settle it

Run the identical GP+CoT protocol on the full BLiMP (67 paradigms), SLING (38), and RuBLiMP (45) benchmarks, averaging over all paradigms with the same three trials per pair. If the small-model gains concentrate in the selected hard categories and the full-benchmark LLM-SLM gap remains near the 13-point baseline (or the 56% reduction shrinks to insignificance), the central claim — that one rule explanation closes most of the capacity gap — would be an artifact of paradigm selection, not a property of grammar prompting.

Watch

Extended reading notes

Core claim

The paper's discovery is that the failure of language models to judge grammatical acceptability despite being able to explain grammar reflects a knowing-versus-using gap, and that the gap can be bridged with the model's own metalinguistic output. The method, grammar prompting, first elicits a beginner-oriented explanation of a target phenomenon (e.g., negative polarity licensing, aspect marking) from a large model, then feeds that explanation back as context to the deciding model before it chooses between two minimal-pair sentences. Across English, Chinese, and Russian, this consistently improves accuracy over base and chain-of-thought conditions. The key quantitative claim is that with a single LLM-generated rule explanation plus chain-of-thought, three smaller models (GPT-3.5, Claude Haiku, and Llama 3.3 9B) finish on average within 5.8 percentage points of GPT-4o and Claude Sonnet, reducing the average LLM-SLM gap from 13.0 to 5.8 percentage points (a 56% relative reduction) on the selected challenging paradigm categories.

Load-bearing premise

The gap-reduction numbers are computed only over manually selected 'challenging' paradigm categories — BLiMP categories where gpt-4o scored at or below 90%, six SLING categories chosen by hand, and RuBLiMP categories at or below 96% — and the paper assumes that subset represents the general LLM-SLM gap.

Editorial extensions

If this is right

  • Small models can be brought within a few points of frontier-LLM accuracy on grammatical acceptability judgments using only a short generated rule explanation, with no fine-tuning.
  • Grammar prompting works best for phenomena with clear distributional constraints or functional morphemes (NPI licensing, aspect, alternative questions) and helps least for constituent-sensitive island effects and lexical classifier-noun agreement.
  • A single relevant explanation is worth more than a compiled set of explanations: the "textbook" condition performed like the irrelevant control, implying models cannot reliably select the right rule from a larger body of grammar.
  • Because explanations are self-generated by the model rather than drawn from curated resources, the paradigm is portable to languages and domains where no grammar textbook exists, though the paper explicitly leaves low-resource languages untested.
  • Combined with chain-of-thought, the LLM-generated explanation lets SLMs reach roughly 90% accuracy or higher in Chinese and Russian where baselines were 74–82%.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • An implied extension, not tested in the paper, is that the same explain-then-process loop could improve other structural tasks where models know a rule but fail to apply it, such as syntax-aware translation or code generation.
  • The 'meaning-first' failure mode identified here suggests the benefit of grammar prompting may be partly to block paraphrase-based shortcuts; a direct testable corollary is that removing the explanation should revert accuracy to baseline, which the control condition already supports.
  • A reader should be cautious that the 13.0 to 5.8 percentage-point gap figure relies on hard paradigm subsets only; a fair comparison on all paradigms of each benchmark would reveal whether the method's average benefit is smaller than the headline number.
  • For languages where models cannot generate reliable explanations, one could test translating grammar prompts from high-resource languages or using a multilingual explanation generated by a frontier model.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. The paper proposes 'grammar prompting' (GP), an explain-then-process paradigm in which a large LLM first generates a concise metalinguistic explanation of a syntactic paradigm, and this explanation is then fed back as context to the target model (an LLM or a smaller SLM) before it chooses the grammatical sentence in a minimal pair. The method is evaluated on English BLiMP, Chinese SLING, and Russian RuBLiMP with five models (GPT-4o, GPT-3.5, Claude 3.5 Sonnet, Claude Haiku, Llama 3.3 9B), across base, CoT, control, textbook, and three-shot conditions. The central reported result is that on SLMs, GP alone reduces the average LLM-SLM accuracy gap by about 20%, and GP+CoT reduces it by 56% (13.0 pp to 5.8 pp), with the abstract and conclusion presenting this as bringing small models close to frontier-LLM performance.

Significance. If the central claim were established on full benchmarks, this would be a practically useful and inexpensive finding: the method requires only a single generated explanation per paradigm, works across three typologically different languages, and the experimental design includes valuable controls (an irrelevant-explanation control, a textbook-style condition, and three-shot baselines). The paper also documents concrete failure cases, which is a strength. However, the headline quantitative claim is computed on a post-hoc subset of 'challenging' paradigms, not on the full benchmarks, and the cross-language average hides substantial heterogeneity. The contribution is therefore better characterized as a promising prompt-engineering result on a selected subset than as the general equalizer claimed in the abstract.

major comments (4)
  1. [§4.2, §5.7, Tables 7–8] The headline '56% reduction' (13.0 pp to 5.8 pp) is computed only over a post-hoc selection of challenging paradigms, not over the full benchmarks. The English gap on the full BLiMP (Table 8: gpt-3.5 84.0 vs. gpt-4o 91.2) is 7.2 pp, whereas the selected English subset in Table 7 has a base gap of 12.9 pp. Because the selection thresholds are based on gpt-4o performance (BLiMP categories at or below 90%, SLING categories excluding easier ones, RuBLiMP categories at or below 96%), the baseline gap is inflated and the general claim that grammar prompting 'cuts the capacity gap by 56%' is not supported for the full benchmarks. The authors should either report the gap analysis on all paradigms or explicitly and prominently reframe the claim as applying to the selected challenging subset.
  2. [Table 7, §5.7] The cross-language average conceals strong per-language heterogeneity. For English, GP alone increases the LLM-SLM gap from 12.9 to 17.4 pp, and GP+CoT leaves it at 9.4 pp, while Chinese and Russian gaps under GP+CoT drop to 3.4 and 4.4 pp. The statement in §5.7 and §7 that three SLMs finish within 5.8 pp 'across English, Chinese, and Russian' is therefore misleading: the 5.8 pp figure is an average, and English remains roughly twice that. The paper should present the per-language gap results as a central result and temper the equalizer claim accordingly.
  3. [§4.2–§4.3, §5.7] No uncertainty quantification is provided for the main effect. Accuracies are averages over the first 50 items per paradigm and three A/B order trials, so item-level bootstrap confidence intervals or per-paradigm standard errors are feasible. Without them, the precise 13.0-to-5.8 pp reduction and the derived 20% and 56% relative reductions cannot be distinguished from sampling noise, especially given only three SLMs and small numbers of selected categories per language.
  4. [§5.3, Tables 3–4, Appendix C.2] The claim that grammar prompting 'consistently outperforms' base and CoT conditions is contradicted by systematic regressions: gpt-3.5 English ellipsis drops from 76.0 (base) to 72.0 (GP+CoT, Sonnet prompt); gpt-4o English ellipsis falls from 70.3 to 59.3 (GP+CoT, Sonnet); and gpt-3.5 Chinese wh-fronting falls from 98.3 to 91.7/88.7 under GP+CoT, a case the paper itself documents in Appendix C.2. The discussion should acknowledge these paradigms and state the scope of the improvement claim rather than presenting the gains as uniform.
minor comments (4)
  1. [§3] There is a typo in 'aninstruction template' near the description of the instruction template; it should read 'an instruction template.'
  2. [§5.3] 'The grammar prompting conditions consistently outperforms basic and chain-of-thought conditions' has a subject-verb agreement error and should be rephrased.
  3. [Table 6] The notation 'GP(b)-o1' is used without an explicit caption definition; the caption should state that this is GP with the o1-generated beginner prompt.
  4. [§4.2] The choice to use the first 50 minimal pairs per paradigm is not justified; if item ordering affects difficulty, this could bias the per-paradigm estimates, so a sentence on why this truncation is appropriate (or a robustness check) would help.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper is an empirical prompting study whose claimed gains are measured, not derived, and whose control conditions rule out the main alternative explanations.

full rationale

This is an empirical prompting study, not a derivation, so the standard circularity failure modes do not apply. The grammar explanations are generated by large LLMs (Claude 3.5 Sonnet and GPT-o1) from an instruction template that names the paradigm, but the explanations are produced without access to the test labels or the specific minimal pairs scored in the evaluation. The target models are then evaluated on held-out minimal pairs from the same paradigm, which is the intended explain-then-process intervention rather than a circular construction. The paper includes a control condition that supplies an irrelevant grammar explanation (the 'null quotative') and reports consistently worse performance than relevant grammar prompts, directly supporting the claim that the content of the explanation drives the gains rather than the mere presence of extra instruction text. The textbook condition similarly shows that simply providing many explanations does not reproduce the benefit, further separating the effect from generic prompt enrichment. No parameter is fitted to the reported accuracies, no uniqueness theorem is imported from the authors' prior work, and no target result is defined in terms of the input. The central quantitative claim, the 56% gap reduction, is computed on a deliberately selected subset of challenging paradigms rather than on the full benchmarks, but this is a generalizability and external-validity concern, not a circularity concern: the paper transparently states the selection criterion (Section 4.2) and reports the full-paradigm results in Appendix B, so the claim is not hidden or constructed to equal its inputs. The absence of significance tests and the observed regressions on some paradigms are correctness risks, but they do not make any claim equivalent to its own inputs by construction. There are no self-citations used as load-bearing support, and the related work citations are independent of the paper's conclusions. The honest finding is therefore no significant circularity.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The central experimental results depend on several hand-chosen design quantities that are not fitted model parameters in the usual sense, but they shape every headline number. The most important is the selection of challenging paradigms. No new entities are introduced.

free parameters (4)
  • Challenging-paradigm selection thresholds = BLiMP <=90%, SLING manual exclusions, RuBLiMP <=96%
    Section 4.2; all averages and the 13.0 to 5.8 pp gap figures in Section 5.7 are computed over this subset, so the thresholds shape the headline result.
  • Per-paradigm item count = 50
    Section 4.2 uses the first 50 minimal pairs per paradigm; no sensitivity analysis shows whether results hold on other subsamples.
  • Grammar prompt audience = beginner (GPb)
    Chosen after comparing beginner vs expert on 18 categories (Section 5.1); all later experiments use GPb, so the main results are conditional on this choice.
  • Explanation source model = Claude 3.5 Sonnet and GPT-o1; gpt-4o replaced by o1 mid-study
    Section 4.4; prompt quality and results vary by generator, but the main gap analysis is aggregated over both sources.
assumptions (4)
  • domain assumption Benchmark labels (BLiMP, SLING, RuBLiMP) are correct grammaticality judgments.
    The evaluation treats dataset labels as ground truth for acceptability, which is standard for these benchmarks but not independently verified in this paper.
  • domain assumption LLM-generated grammar explanations are accurate and relevant to the target paradigm.
    The paper relies on only 'small-scale manual verification' (Section 4.4) and does not systematically audit explanation correctness.
  • domain assumption English prompts transfer effectively to Chinese and Russian minimal-pair judgments.
    All prompts are in English while target sentences are in Chinese or Russian; the authors cite Wendler et al. but do not compare prompt-language variants.
  • domain assumption Models actually use the provided explanation when making their choice.
    The mechanism claim assumes the explanation influences the judgment; the control condition is partial evidence, but no causal probe is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Explain-then-Process: Using Grammar Prompting to Enhance Grammatical Acceptability Judgments." pith.science (2026). https://pith.science/paper/MHJOFBXD

@misc{pith2026250602302,
  author       = {Pith},
  title        = {Pith review of: Explain-then-Process: Using Grammar Prompting to Enhance Grammatical Acceptability Judgments},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MHJOFBXD}},
  note         = {Machine review of arXiv:2506.02302}
}
read the original abstract

Large language models (LLMs) can explain grammatical rules, yet they often fail to apply those rules when judging sentence acceptability. We present "grammar prompting", an explain-then-process paradigm: a large LLM first produces a concise explanation of the relevant syntactic phenomenon, then that explanation is fed back as additional context to the target model -- either an LLM or a smaller language model (SLM) -- before deciding which sentence of a minimal pair is grammatical. On the English BLiMP, Chinese SLING, and Russian RuBLiMP benchmarks, this simple prompt design yields substantial improvements over strong baselines across many syntactic phenomena. Feeding an LLM's metalinguistic explanation back to the target model bridges the gap between knowing a rule and using it. On SLMs, grammar prompting alone trims the average LLM-SLM accuracy gap by about 20%, and when paired with chain-of-thought, by 56% (13.0 pp -> 5.8 pp), all at negligible cost. The lightweight, language-agnostic cue lets low-cost SLMs approach frontier-LLM performance in multilingual settings.

Figures

Figures reproduced from arXiv: 2506.02302 by the authors.

Figure 1
Figure 1. An illustration of Chain-of-Thought (CoT) reasoning vs. Grammar Prompting (GP) vs. Grammar [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Instruction Template for Eliciting Grammar [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Sonnet-generated ‘Beginner-Oriented’ Gram [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Instruction for generating grammar explanations. [PITH_FULL_IMAGE:figures/full_fig_p013_4.png]
Figure 5
Figure 5. Figure 5: Sonnet-generated Expert grammar prompt for ‘left branch island echo question’ in the BLiMP paradigm. [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]
Figure 6
Figure 6. Figure 6: Sonnet and o1-generated Beginner grammar prompt for ‘left branch island echo question’ in the BLiMP [PITH_FULL_IMAGE:figures/full_fig_p014_6.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. The Gold Medals in an Empty Room: Diagnosing Metalinguistic Reasoning in LLMs with Camlang

    cs.CL 2025-08 conditional novelty 7.0 of 10

    A novel constructed language with explicit grammar and dictionary reveals a large gap between human metalinguistic learning (87%) and the best LLM (47%) on translated CommonsenseQA.

Reference graph

Works this paper leans on

23 extracted references · 7 canonical work pages · cited by 1 Pith paper

  1. [1]

    online" 'onlinestring :=

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint eprinttype howpublished institution journal key month note number organization pages publisher school series title type volume year doi pubmed url lastchecked label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block STRING...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    Wafa Abdullah Alrajhi, Hend Al-Khalifa, and Abdulmalik AlSalman. 2022. https://doi.org/10.18653/v1/2022.wanlp-1.17 Assessing the linguistic knowledge in A rabic pre-trained language models using minimal pairs . In Proceedings of the Seventh Arabic Natural Language Processing Workshop (WANLP), pages 185--193, Abu Dhabi, United Arab Emirates (Hybrid). Assoc...

    work page doi:10.18653/v1/2022.wanlp-1.17 2022

  4. [4]

    Seth Aycock, David Stap, Di Wu, Christof Monz, and Khalil Sima'an. 2024. https://arxiv.org/abs/2409.19151 Can llms really learn to translate a low-resource language from one grammar book? Preprint, arXiv:2409.19151

    work page Pith review arXiv 2024

  5. [5]

    Shabnam Behzad, Amir Zeldes, and Nathan Schneider. 2024. https://doi.org/10.18653/v1/2024.findings-emnlp.916 To ask LLM s about E nglish grammaticality, prompt them in a different language . In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 15622--15634, Miami, Florida, USA. Association for Computational Linguistics

    doi:10.18653/v1/2024.findings-emnlp.916 2024

  6. [6]

    Vittoria Dentella, Fritz Günther, and Evelina Leivada. 2023. https://doi.org/10.1073/pnas.2309583120 Systematic testing of three language models reveals low language accuracy, absence of response stability, and a yes-response bias . Proceedings of the National Academy of Sciences, 120(51):e2309583120

    doi:10.1073/pnas.2309583120 2023

  7. [7]

    Mengnan Du, Fengxiang He, Na Zou, Dacheng Tao, and Xia Hu. 2023. Shortcut learning of large language models in natural language understanding. Communications of the ACM, 67(1):110--120

    2023

  8. [8]

    Ping Guo, Yubing Ren, Yue Hu, Yunpeng Li, Jiarui Zhang, Xingsheng Zhang, and Heyan Huang. 2024. https://aclanthology.org/2024.lrec-main.1362 Teaching large language models to translate on low-resource languages with textbook prompting . In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluatio...

    work page 2024

Show all 23 references
  1. [9]

    Andrew Lampinen, Ishita Dasgupta, Stephanie Chan, Kory Mathewson, Mh Tessler, Antonia Creswell, James McClelland, Jane Wang, and Felix Hill. 2022. https://doi.org/10.18653/v1/2022.findings-emnlp.38 Can language models learn from explanations in context? In Findings of the Asso...

    2022 doi

  2. [10]

    Wei Qi Leong, Jian Gang Ngui, Yosephine Susanto, Hamsawardhini Rengarajan, Kengatharaiyer Sarveswaran, and William Chandra Tjhi. 2023. https://arxiv.org/abs/2309.06085 Bhasa: A holistic southeast asian linguistic and cultural evaluation suite for large language models . Prepri...

    2023 arXiv

  3. [11]

    Ivanova, Idan A

    Kyle Mahowald, Anna A. Ivanova, Idan A. Blank, Nancy Kanwisher, Joshua B. Tenenbaum, and Evelina Fedorenko. 2024. https://arxiv.org/abs/2301.06627 Dissociating language and thought in large language models . Preprint, arXiv:2301.06627

    2024 arXiv

  4. [12]

    Zhuang Qiu, Xufeng Duan, and Zhenguang G. Cai. 2024. https://arxiv.org/abs/2406.11116 Grammaticality representation in chatgpt as compared to linguists and laypeople . Preprint, arXiv:2406.11116

    2024 arXiv

  5. [13]

    Carson Sch \"u tze. 2016. The empirical base of linguistics: Grammaticality judgments and linguistic methodology. Language Science Press

    2016

  6. [14]

    Yixiao Song, Kalpesh Krishna, Rajesh Bhatt, and Mohit Iyyer. 2022. https://doi.org/10.18653/v1/2022.emnlp-main.305 SLING : S ino linguistic evaluation of large language models . In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 46...

    2022 doi

  7. [15]

    Brown, Adam Santoro, Aditya Gupta, Adrià Garriga-Alonso, Agnieszka Kluska, Aitor Lewkowycz, Akshat Agarwal, Alethea Power, Alex Ray, Alex Warstadt, Alexander W

    Aarohi Srivastava, Abhinav Rastogi, Abhishek Rao, Abu Awal Md Shoeb, Abubakar Abid, Adam Fisch, Adam R. Brown, Adam Santoro, Aditya Gupta, Adrià Garriga-Alonso, Agnieszka Kluska, Aitor Lewkowycz, Akshat Agarwal, Alethea Power, Alex Ray, Alex Warstadt, Alexander W. Kocurek, Ali...

    2023 arXiv

  8. [16]

    Michelle Suijkerbuijk, Zo \"e Prins, Marianne de Heer Kloots, Willem Zuidema, and Stefan L Frank. 2025. Blimp-nl: A corpus of dutch minimal pairs and acceptability judgments for language model evaluation. Computational Linguistics, pages 1--39

    2025

  9. [17]

    Ekaterina Taktasheva, Maxim Bazhukov, Kirill Koncha, Alena Fenogenova, Ekaterina Artemova, and Vladislav Mikhailov. 2024. https://arxiv.org/abs/2406.19232 Rublimp: Russian benchmark of linguistic minimal pairs . Preprint, arXiv:2406.19232

    2024 arXiv

  10. [18]

    Garrett Tanzer, Mirac Suzgun, Eline Visser, Dan Jurafsky, and Luke Melas-Kyriazi. 2024. https://openreview.net/forum?id=tbVWug9f2h A benchmark for learning to translate a new language from one grammar book . In The Twelfth International Conference on Learning Representations

    2024

  11. [19]

    Alex Warstadt, Alicia Parrish, Haokun Liu, Anhad Mohananey, Wei Peng, Sheng-Fu Wang, and Samuel R. Bowman. 2020. https://doi.org/10.1162/tacl_a_00321 BL i MP : The benchmark of linguistic minimal pairs for E nglish . Transactions of the Association for Computational Linguistic...

    2020 doi

  12. [20]

    Alex Warstadt, Amanpreet Singh, and Samuel R. Bowman. 2019. https://doi.org/10.1162/tacl_a_00290 Neural Network Acceptability Judgments . Transactions of the Association for Computational Linguistics, 7:625--641

    2019 doi

  13. [21]

    Chris Wendler, Veniamin Veselovsky, Giovanni Monea, and Robert West. 2024. https://doi.org/10.18653/v1/2024.acl-long.820 Do llamas work in E nglish? on the latent language of multilingual transformers . In Proceedings of the 62nd Annual Meeting of the Association for Computati...

    2024 doi

  14. [22]

    Noam Wies, Yoav Levine, and Amnon Shashua. 2024. The learnability of in-context learning. Advances in Neural Information Processing Systems, 36

    2024

  15. [23]

    Chen Zhang, Xiao Liu, Jiuheng Lin, and Yansong Feng. 2024. https://doi.org/10.18653/v1/2024.findings-acl.519 Teaching large language models an unseen language on the fly . In Findings of the Association for Computational Linguistics ACL 2024, pages 8783--8800, Bangkok, Thailan...

    2024 doi

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.