mlsanityData to 7 Oct 2026

Different types of syntactic agreement recruit the same units within large language models

arXiv 2512.03676 ↗First posted 3 Dec 2025cs.CLcited by 7

Not on the map. Why?

Abstract (arXiv)

Large language models (LLMs) can reliably distinguish grammatical from ungrammatical sentences, but how grammatical knowledge is represented within the models remains an open question. We investigate whether different syntactic phenomena recruit shared or distinct components in LLMs. Using a functional localization approach inspired by cognitive neuroscience, we identify the LLM units most responsive to 67 English syntactic phenomena in seven open-weight models. These units are consistently recruited across sentences containing the phenomena and causally support the models' syntactic performance. Critically, different types of syntactic agreement (e.g., subject-verb, anaphor, determiner-noun) recruit overlapping sets of units, suggesting that agreement constitutes a meaningful functional category for LLMs. This pattern holds in English, Russian, and Chinese; and further, in a cross-lingual analysis of 57 diverse languages, structurally more similar languages share more units for subject-verb agreement. Taken together, these findings reveal that syntactic agreement-a critical marker of syntactic dependencies-constitutes a meaningful category within LLMs' representational spaces.

Key facts. Cited by 7 papers in the map (data through 7 Oct 2026). Builds on 40 of its 71 references that are in the map. First posted on arXiv 3 Dec 2025.

Builds on

40 of 71 references.

References · 40

  1. 2019HuggingFace's Transformers: State-of-the-art Natural Language Processingcited by 2,201
  2. 2019BERT Rediscovers the Classical NLP Pipelinecited by 1,046
  3. 2019Right for the Wrong Reasons: Diagnosing Syntactic Heuristics in Natural Language Inferencecited by 864
  4. 2019BLiMP: The Benchmark of Linguistic Minimal Pairs for Englishcited by 553
  5. 2025Phi-4-Mini Technical Report: Compact yet Powerful Multimodal Language Models via Mixture-of-LoRAscited by 491
  6. 2024Sparse Feature Circuits: Discovering and Editing Interpretable Causal Graphs in Language Modelscited by 412
  7. 2016Assessing the Ability of LSTMs to Learn Syntax-Sensitive Dependenciescited by 350
  8. 2018Targeted Syntactic Evaluation of Language Modelscited by 239
  9. 2018Colorless green recurrent networks dream hierarchicallycited by 210
  10. 2023Finding Alignments Between Interpretable Causal Variables and Distributed Neural Representationscited by 197
  11. 2019Neural Language Models as Psycholinguistic Subjects: Representations of Syntactic Statecited by 137
  12. 2020Finding Universal Grammatical Relations in Multilingual BERTcited by 108
28 more
  1. 2020Analyzing Individual Neurons in Pre-trained Language Modelscited by 97
  2. 2019The emergence of number and syntax units in LSTM language modelscited by 92
  3. 2018What do RNN Language Models Learn about Filler-Gap Dependencies?cited by 86
  4. 2025MultiBLiMP 1.0: A Massively Multilingual Benchmark of Linguistic Minimal Pairscited by 65
  5. 2019Investigating BERT's Knowledge of Language: Five Analysis Methods with NPIscited by 58
  6. 2022minicons: Enabling Flexible Behavioral and Representational Analyses of Transformer Language Modelscited by 58
  7. 2025Large Language Models Share Representations of Latent Grammatical Concepts Across Typologically Diverse Languagescited by 57
  8. 2018Evaluating Layers of Representation in Neural Machine Translation on Part-of-Speech and Semantic Tagging Taskscited by 56
  9. 2025Large Language Models Pass the Turing Testcited by 48
  10. 2025Lost in Multilinguality: Dissecting Cross-lingual Factual Inconsistency in Transformer Language Modelscited by 43
  11. 2020Mechanisms for Handling Nested Dependencies in Neural-Network Language Models and Humanscited by 42
  12. 2021Deep Subjecthood: Higher-Order Grammatical Features in Multilingual BERTcited by 41
  13. 2023Does GPT-4 pass the Turing test?cited by 37
  14. 2024The LLM Language Network: A Neuroscientific Approach for Identifying Causally Task-Relevant Unitscited by 34
  15. 2024The Same But Different: Structural Similarities and Differences in Multilingual Language Modelingcited by 27
  16. 2024On the Similarity of Circuits across Languages: a Case Study on the Subject-verb Agreement Taskcited by 24
  17. 2024RuBLiMP: Russian Benchmark of Linguistic Minimal Pairscited by 23
  18. 2021Disentangling Semantics and Syntax in Sentence Embeddings with Pre-trained Language Modelscited by 22
  19. 2025Causal Interventions Reveal Shared Structure Across English Filler-Gap Constructionscited by 21
  20. 2025Paths Not Taken: Understanding and Mending the Multilingual Factual Recall Pipelinecited by 20
  21. 2023Structural Priming Demonstrates Abstract Grammatical Representations in Multilingual Language Modelscited by 16
  22. 2022SLING: Sino Linguistic Evaluation of Large Language Modelscited by 15
  23. 2023Uncovering Intermediate Variables in Transformers using Circuit Probingcited by 12
  24. 2023Probing LLMs for Joint Encoding of Linguistic Categoriescited by 9
  25. 2025TurBLiMP: A Turkish Benchmark of Linguistic Minimal Pairscited by 9
  26. 2025On the Acquisition of Shared Grammatical Representations in Bilingual Language Modelscited by 7
  27. 2025How Syntax Specialization Emerges in Language Modelscited by 7
  28. 2019Representation of Constituents in Neural Language Models: Coordination Phrase as a Case Studycited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 7.

7 papers.

  1. 2026Modular Cognitive Architecture Emerges in Large Language Modelscited by 2
  2. 2026Filling in the Mechanisms: How do LMs Learn Filler-Gap Dependencies under Developmental Constraints?cited by 0
  3. 2026Fine-Grained Analysis of Shared Syntactic Mechanisms in Language Modelscited by 0
  4. 2026Linear representations of grammaticality in neural language modelscited by 0
  5. 2026Cognitive Convergence: Deep Similarities Between Large Language Models and Human Cognitioncited by 0
  6. 2026Focus particles and scalar inferences across humans and language modelscited by 0
  7. 2026Assessing the Impact of Language Disparity on Multilingual Linguistic Ability in Large Language Modelscited by 0