# How Syntax Specialization Emerges in Language Models

arXiv: https://arxiv.org/abs/2505.19548 · HTML: https://mlsanity.com/paper/2505.19548/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- Builds on 20 of its 36 references that are in the map.
- First posted on arXiv 26 May 2025.

## Builds on (most cited first, 20 of 20)

- [Pythia: A Suite for Analyzing Large Language Models Across Training and Scaling](https://mlsanity.com/paper/2304.01373/) (2023)
- [Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 small](https://mlsanity.com/paper/2211.00593/) (2022)
- [What Does BERT Look At? An Analysis of BERT's Attention](https://mlsanity.com/paper/1906.04341/) (2019)
- [In-context Learning and Induction Heads](https://mlsanity.com/paper/2209.11895/) (2022)
- [Analyzing Multi-Head Self-Attention: Specialized Heads Do the Heavy Lifting, the Rest Can Be Pruned](https://mlsanity.com/paper/1905.09418/) (2019)
- [BLiMP: The Benchmark of Linguistic Minimal Pairs for English](https://mlsanity.com/paper/1912.00582/) (2019)
- [Assessing the Ability of LSTMs to Learn Syntax-Sensitive Dependencies](https://mlsanity.com/paper/1611.01368/) (2016)
- [Targeted Syntactic Evaluation of Language Models](https://mlsanity.com/paper/1808.09031/) (2018)
- [Language-Specific Neurons: The Key to Multilingual Capabilities in Large Language Models](https://mlsanity.com/paper/2402.16438/) (2024)
- [Sudden Drops in the Loss: Syntax Acquisition, Phase Transitions, and Simplicity Bias in MLMs](https://mlsanity.com/paper/2309.07311/) (2023)
- [The emergence of number and syntax units in LSTM language models](https://mlsanity.com/paper/1903.07435/) (2019)
- [Large Language Models Share Representations of Latent Grammatical Concepts Across Typologically Diverse Languages](https://mlsanity.com/paper/2501.06346/) (2025)
- [Language models align with human judgments on key grammatical constructions](https://mlsanity.com/paper/2402.01676/) (2024)
- [PolyPythias: Stability and Outliers across Fifty Language Model Pre-Training Runs](https://mlsanity.com/paper/2503.09543/) (2025)
- [From Language to Cognition: How LLMs Outgrow the Human Language Network](https://mlsanity.com/paper/2503.01830/) (2025)
- [Converging to a Lingua Franca: Evolution of Linguistic Regions and Semantics Alignment in Multilingual Large Language Models](https://mlsanity.com/paper/2410.11718/) (2024)
- [Causal Analysis of Syntactic Agreement Neurons in Multilingual Language Models](https://mlsanity.com/paper/2210.14328/) (2022)
- [Unveiling Language Competence Neurons: A Psycholinguistic Approach to Model Interpretability](https://mlsanity.com/paper/2409.15827/) (2024)
- [Grammaticality Representation in ChatGPT as Compared to Linguists and Laypeople](https://mlsanity.com/paper/2406.11116/) (2024)
- [Brain-Inspired Exploration of Functional Networks and Key Neurons in Large Language Models](https://mlsanity.com/paper/2502.20408/) (2025)

## Built on by (most cited first, 7 of 7)

- [Different types of syntactic agreement recruit the same units within large language models](https://mlsanity.com/paper/2512.03676/) (2025)
- [Negative Pre-activations Differentiate Syntax](https://mlsanity.com/paper/2509.24198/) (2025)
- [A Review of Developmental Interpretability in Large Language Models](https://mlsanity.com/paper/2508.15841/) (2025)
- [Discovering Semantic Subdimensions through Disentangled Conceptual Representations](https://mlsanity.com/paper/2508.21436/) (2025)
- [The Grammar of Transformers: A Systematic Review of Interpretability Research on Syntactic Knowledge in Language Models](https://mlsanity.com/paper/2601.19926/) (2026)
- [Linear representations of grammaticality in neural language models](https://mlsanity.com/paper/2607.15175/) (2026)
- [Do Large Language Models Grasp The Grammar? Evidence from Grammar-Book-Guided Probing in Luxembourgish](https://mlsanity.com/paper/2510.24856/) (2025)

## Abstract (arXiv)

Large language models (LLMs) have been found to develop surprising internal specializations: Individual neurons, attention heads, and circuits become selectively sensitive to syntactic structure, reflecting patterns observed in the human brain. While this specialization is well-documented, how it emerges during training and what influences its development remains largely unknown. In this work, we tap into the black box of specialization by tracking its formation over time. By quantifying internal syntactic consistency across minimal pairs from various syntactic phenomena, we identify a clear developmental trajectory: Syntactic sensitivity emerges gradually, concentrates in specific layers, and exhibits a 'critical period' of rapid internal specialization. This process is consistent across architectures and initialization parameters (e.g., random seeds), and is influenced by model scale and training data. We therefore reveal not only where syntax arises in LLMs but also how some models internalize it during training. To support future research, we will release the code, models, and training checkpoints upon acceptance.