--- title: "Developmentally-plausible Working Memory Shapes a Critical Period for Language — Mita 2025" description: "Drawing on the Critical Period Hypothesis and Newport's Less-is-More Hypothesis, the authors test whether mimicking… 2 references, cited by 7 papers here." authors: ["Masato Mita", "Ryo Yoshida", "Yohei Oseki"] year: 2025 venue: "Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)" source: "https://aclanthology.org/2025.acl-long.462/" doi: "10.18653/v1/2025.acl-long.462" licence: "CC BY 4.0 (ACL Anthology)" licence_url: "https://creativecommons.org/licenses/by/4.0/" summary_licence: "Elamur Cosmic License" canonical: "https://research.elamur.ai/free/en/papers/mita-2025-developmentally-plausible-working-memory-shapes/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/mita-2025-developmentally-plausible-working-memory-shapes/) [home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2025](https://md.research.elamur.ai/free/en/papers/2025/page/2/) / Developmentally-plausible Working Memory Shapes a Critical Period for Language Acquisition # Developmentally-plausible Working Memory Shapes a Critical Period for Language Acquisition **Authors:** [Masato Mita](https://md.research.elamur.ai/free/en/authors/masato-mita/), [Ryo Yoshida](https://md.research.elamur.ai/free/en/authors/ryo-yoshida/) and [Yohei Oseki](https://md.research.elamur.ai/free/en/authors/yohei-oseki/) · **Year:** 2025 · **Venue:** Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/) **Links:** [Publisher's page](https://aclanthology.org/2025.acl-long.462/) · [PDF](https://aclanthology.org/2025.acl-long.462.pdf) · [DOI 10.18653/v1/2025.acl-long.462](https://doi.org/10.18653/v1/2025.acl-long.462) · [OpenAlex](https://openalex.org/W4412945478) **Citations:** 2 citations (OpenAlex, 2026-10-14) · [7](https://md.research.elamur.ai/free/en/papers/mita-2025-developmentally-plausible-working-memory-shapes/cited-by/) in our library, of the 140,884 papers whose bibliographies we read ## Summary LLMs need far more training data than children to reach comparable linguistic competence. Drawing on the Critical Period Hypothesis and Newport's Less-is-More Hypothesis, the authors test whether mimicking how children's working memory grows can make LM training more data-efficient. They model working-memory growth across development as an exponential curve of the form y = b − aˣ (0 \< a \< 1), which rises steeply early and then levels off at an adult ceiling. They turn this curve into a training schedule for GPT-2 called DynamicLimit-Exp. Memory is tightly restricted at the start of training and the restriction is loosened exponentially over time. The model is trained on the child-directed speech corpus AO-CHILDES and also on Wikipedia. It is evaluated with Zorro, a targeted syntactic benchmark designed for child-directed speech. DynamicLimit-Exp beats both an unconstrained baseline and a fixed-limit baseline. The result points to a cognitively motivated inductive bias for data-efficient LMs. It also gives indirect, computational-level support for working-memory development as a mechanism behind the critical period. ## Key points - The schedule is grounded in three developmental stages of human working memory: 2–7 years (rapid growth), 8–14 years (slowing growth) and 15+ years (adult plateau). The authors argue that linear or logarithmic curves cannot capture both the fast early rise and the eventual ceiling. - The baselines are NoLimit (no memory constraint) and a static memory limit. DynamicLimit-Exp overtakes both by epoch 10 in several Zorro categories, with the clearest late-stage gains on CASE and FILLER.GAP. - In embedding analyses using t-SNE, entropy and inter-cluster distance, NoLimit's representations collapse after epoch 5. DynamicLimit-Exp keeps embeddings more dispersed and continues to separate them across epochs. - A sentence-length test used four Wikipedia sets of 500,000 sentences each. DynamicLimit-Exp was significantly better for lengths 11–50 (58.7 vs 47.0), 51–100 (42.5 vs 40.6) and 101–150 (40.8 vs 37.3). NoLimit was marginally ahead only on 5–10-word sentences (47.2 vs 46.8). - The code is released as CPLM at github.com/osekilab/CPLM. The paper appeared at ACL 2025 (Long Papers). *Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.* Fields: [Natural language processing](https://md.research.elamur.ai/free/en/shelves/nlp/) Topics: [🧒 Psycholinguistics & language acquisition](https://md.research.elamur.ai/free/en/topics/psycholinguistics/) ## Cite Cite as: Masato Mita, Ryo Yoshida and Yohei Oseki. 2025. Developmentally-plausible Working Memory Shapes a Critical Period for Language Acquisition. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). The paper itself, not this page. BibTeX from the library's record of the work. ```bibtex @inproceedings{mita2025developmentally, title = {{Developmentally-plausible Working Memory Shapes a Critical Period for Language Acquisition}}, author = {Masato Mita, Ryo Yoshida and Yohei Oseki}, booktitle = {Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)}, year = {2025}, doi = {10.18653/v1/2025.acl-long.462}, url = {https://doi.org/10.18653/v1/2025.acl-long.462} } ``` - [← previous Child-Directed Language Does Not Consistently Boost Syntax Learning in Language Models](https://md.research.elamur.ai/free/en/papers/padovani-2025-child-directed-language-does-not/) - [next → TriFine: A Large-Scale Dataset of Vision-Audio-Subtitle for Tri-Modal Machine Translation and Benchmark with Fine-Grained Annotated Tags](https://md.research.elamur.ai/free/en/papers/guan-2025-trifine-large-scale-dataset-vision/) - [↑ back to papers of 2025](https://md.research.elamur.ai/free/en/papers/2025/page/2/) ## References 2 references, 0 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet. 1. We thank the anonymous reviewers for their helpful comments and suggestions. We are also grateful to Akiyo Fukatsu for her insightful comments and the Computational Models of Language (CoLa) Reading Group at UC Irvine for their constructive feedback. This work was supported by JSPS KAK-ENHI Grant Nu… [find](https://scholar.google.com/scholar?q=We+thank+the+anonymous+reviewers+for+their+helpful+comments+and+suggestions%2E+We+are+also+grateful+to+Akiyo+Fukatsu+for+her+insightful+comments+and+the+Computational+Models+of+Language+%28CoLa%29+Reading+G) 2. John R. Anderson and Robert Milson. 1989. Human memory: An adaptive perspective. Psychological Review, 96(4):703–719. \#\# Limitations [find](https://scholar.google.com/scholar?q=John+R%2E+Anderson+and+Robert+Milson%2E+1989%2E+Human+memory%3A+An+adaptive+perspective%2E+Psychological+Review%2C+96%284%29%3A703%E2%80%93719%2E+%5C%23%5C%23+Limitations)