# Recurrent babbling: evaluating the acquisition of grammar from limited input data

arXiv: https://arxiv.org/abs/2010.04637 · HTML: https://mlsanity.com/paper/2010.04637/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 9 Oct 2020.

## Built on by (most cited first, 9 of 9)

- [How much do language models copy from their training data? Evaluating linguistic novelty in text generation using RAVEN](https://mlsanity.com/paper/2111.09509/) (2021)
- [What Artificial Neural Networks Can Tell Us About Human Language Acquisition](https://mlsanity.com/paper/2208.07998/) (2022)
- [Findings of the BabyLM Challenge: Sample-Efficient Pretraining on Developmentally Plausible Corpora](https://mlsanity.com/paper/2504.08165/) (2025)
- [How poor is the stimulus? Evaluating hierarchical generalization in neural networks trained on child-directed speech](https://mlsanity.com/paper/2301.11462/) (2023)
- [Can Language Models Be Tricked by Language Illusions? Easier with Syntax, Harder with Semantics](https://mlsanity.com/paper/2311.01386/) (2023)
- [Do Construction Distributions Shape Formal Language Learning In German BabyLMs?](https://mlsanity.com/paper/2503.11593/) (2025)
- [CALaMo: a Constructionist Assessment of Language Models](https://mlsanity.com/paper/2302.03589/) (2023)
- [Age-Dependent Analysis and Stochastic Generation of Child-Directed Speech](https://mlsanity.com/paper/2405.07700/) (2024)
- [Child-directed speech facilitates production, not comprehension, in BabyLMs](https://mlsanity.com/paper/2606.01045/) (2026)

## Abstract (arXiv)

Recurrent Neural Networks (RNNs) have been shown to capture various aspects of syntax from raw linguistic input. In most previous experiments, however, learning happens over unrealistic corpora, which do not reflect the type and amount of data a child would be exposed to. This paper remedies this state of affairs by training a Long Short-Term Memory network (LSTM) over a realistically sized subset of child-directed input. The behaviour of the network is analysed over time using a novel methodology which consists in quantifying the level of grammatical abstraction in the model's generated output (its "babbling"), compared to the language it has been exposed to. We show that the LSTM indeed abstracts new structuresas learning proceeds.