# Beyond Memorization: Assessing Semantic Generalization in Large Language Models Using Phrasal Constructions

arXiv: https://arxiv.org/abs/2501.04661 · HTML: https://mlsanity.com/paper/2501.04661/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Builds on 18 of its 50 references that are in the map.
- First posted on arXiv 8 Jan 2025.

## Builds on (most cited first, 18 of 18)

- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [Scaling Laws for Neural Language Models](https://mlsanity.com/paper/2001.08361/) (2020)
- [A large annotated corpus for learning natural language inference](https://mlsanity.com/paper/1508.05326/) (2015)
- [NLP Evaluation in trouble: On the Need to Measure LLM Data Contamination for each Benchmark](https://mlsanity.com/paper/2310.18018/) (2023)
- [Reasoning or Reciting? Exploring the Capabilities and Limitations of Language Models Through Counterfactual Tasks](https://mlsanity.com/paper/2307.02477/) (2023)
- [Leak, Cheat, Repeat: Data Contamination and Evaluation Malpractices in Closed-Source LLMs](https://mlsanity.com/paper/2402.03927/) (2024)
- [Transformers as Algorithms: Generalization and Stability in In-context Learning](https://mlsanity.com/paper/2301.07067/) (2023)
- [Data Mixing Laws: Optimizing Data Mixtures by Predicting Language Modeling Performance](https://mlsanity.com/paper/2403.16952/) (2024)
- [What and How does In-Context Learning Learn? Bayesian Model Averaging, Parameterization, and Generalization](https://mlsanity.com/paper/2305.19420/) (2023)
- [Are Emergent Abilities in Large Language Models just In-Context Learning?](https://mlsanity.com/paper/2309.01809/) (2023)
- [Language Models Learn Rare Phenomena from Less Rare Phenomena: The Case of the Missing AANNs](https://mlsanity.com/paper/2403.19827/) (2024)
- [How Should Pre-Trained Language Models Be Fine-Tuned Towards Adversarial Robustness?](https://mlsanity.com/paper/2112.11668/) (2021)
- [CxGBERT: BERT meets Construction Grammar](https://mlsanity.com/paper/2011.04134/) (2020)
- [Neural reality of argument structure constructions](https://mlsanity.com/paper/2202.12246/) (2022)
- [A Discerning Several Thousand Judgments: GPT-3 Rates the Article + Adjective + Numeral + Noun Construction](https://mlsanity.com/paper/2301.12564/) (2023)
- [A Method for Studying Semantic Construal in Grammatical Constructions with Interpretable Contextual Embedding Spaces](https://mlsanity.com/paper/2305.18598/) (2023)
- [Constructions Are So Difficult That Even Large Language Models Get Them Right for the Wrong Reasons](https://mlsanity.com/paper/2403.17760/) (2024)
- [UCxn: Typologically Informed Annotation of Constructions Atop Universal Dependencies](https://mlsanity.com/paper/2403.17748/) (2024)

## Built on by (most cited first, 5 of 5)

- [Unpacking Let Alone: Human-Scale Models Generalize to a Rare Construction in Form but not Meaning](https://mlsanity.com/paper/2506.04408/) (2025)
- [Convergent Representations of Linguistic Constructions in Human and Artificial Neural Systems](https://mlsanity.com/paper/2603.29617/) (2026)
- [CxMP: A Linguistic Minimal-Pair Benchmark for Evaluating Constructional Understanding in Language Models](https://mlsanity.com/paper/2602.21978/) (2026)
- [Language Models Learn Constructional Semantics, Not To Mention Syntax: Investigating LM Understanding of Paired-Focus Constructions](https://mlsanity.com/paper/2605.31586/) (2026)
- [Linguistic Productivity in Large Language Models: Models Coerce, but do not Preempt](https://mlsanity.com/paper/2606.02953/) (2026)

## Abstract (arXiv)

The web-scale of pretraining data has created an important evaluation challenge: to disentangle linguistic competence on cases well-represented in pretraining data from generalization to out-of-domain language, specifically the dynamic, real-world instances less common in pretraining data. To this end, we construct a diagnostic evaluation to systematically assess natural language understanding in LLMs by leveraging Construction Grammar (CxG). CxG provides a psycholinguistically grounded framework for testing generalization, as it explicitly links syntactic forms to abstract, non-lexical meanings. Our novel inference evaluation dataset consists of English phrasal constructions, for which speakers are known to be able to abstract over commonplace instantiations in order to understand and produce creative instantiations. Our evaluation dataset uses CxG to evaluate two central questions: first, if models can 'understand' the semantics of sentences for instances that are likely to appear in pretraining data less often, but are intuitive and easy for people to understand. Second, if LLMs can deploy the appropriate constructional semantics given constructions that are syntactically identical but with divergent meanings. Our results demonstrate that state-of-the-art models, including GPT-o1, exhibit a performance drop of over 40% on our second task, revealing a failure to generalize over syntactically identical forms to arrive at distinct constructional meanings in the way humans do. We make our novel dataset and associated experimental data, including prompts and model responses, publicly available.