# Does Vision Accelerate Hierarchical Generalization in Neural Language Learners?

arXiv: https://arxiv.org/abs/2302.00667 · HTML: https://mlsanity.com/paper/2302.00667/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 22 of its 58 references that are in the map.
- First posted on arXiv 1 Feb 2023.

## Builds on (most cited first, 22 of 22)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://mlsanity.com/paper/2103.14030/) (2021)
- [Training data-efficient image transformers & distillation through attention](https://mlsanity.com/paper/2012.12877/) (2020)
- [Zero-Shot Text-to-Image Generation](https://mlsanity.com/paper/2102.12092/) (2021)
- [BEiT: BERT Pre-Training of Image Transformers](https://mlsanity.com/paper/2106.08254/) (2021)
- [RandAugment: Practical automated data augmentation with a reduced search space](https://mlsanity.com/paper/1909.13719/) (2019)
- [Right for the Wrong Reasons: Diagnosing Syntactic Heuristics in Natural Language Inference](https://mlsanity.com/paper/1902.01007/) (2019)
- [BLiMP: The Benchmark of Linguistic Minimal Pairs for English](https://mlsanity.com/paper/1912.00582/) (2019)
- [What Artificial Neural Networks Can Tell Us About Human Language Acquisition](https://mlsanity.com/paper/2208.07998/) (2022)
- [Does syntax need to grow on trees? Sources of hierarchical inductive bias in sequence-to-sequence networks](https://mlsanity.com/paper/2001.03632/) (2020)
- [Cognitive Science in the era of Artificial Intelligence: A roadmap for reverse-engineering the infant language-learner](https://mlsanity.com/paper/1607.08723/) (2016)
- [Visually Grounded Neural Syntax Acquisition](https://mlsanity.com/paper/1906.02890/) (2019)
- [Can neural networks acquire a structural bias from raw linguistic data?](https://mlsanity.com/paper/2007.06761/) (2020)
- [Revisiting the poverty of the stimulus: hierarchical generalization without a hierarchical bias in recurrent neural networks](https://mlsanity.com/paper/1802.09091/) (2018)
- [Coloring the Blank Slate: Pre-training Imparts a Hierarchical Inductive Bias to Sequence-to-sequence Models](https://mlsanity.com/paper/2203.09397/) (2022)
- [Recurrent Neural Networks in Linguistic Theory: Revisiting Pinker and Prince (1988) and the Past Tense Debate](https://mlsanity.com/paper/1807.04783/) (2018)
- [Does Vision-and-Language Pretraining Improve Lexical Grounding?](https://mlsanity.com/paper/2109.10246/) (2021)
- [Compositional Generalization in Image Captioning](https://mlsanity.com/paper/1909.04402/) (2019)
- [What is Learned in Visually Grounded Neural Syntax Acquisition](https://mlsanity.com/paper/2005.01678/) (2020)
- [A computational model of early language acquisition from audiovisual experiences of young infants](https://mlsanity.com/paper/1906.09832/) (2019)

## Built on by (most cited first, 1 of 1)

- [LaViSA: A Language and Vision Structural Ambiguity Benchmark](https://mlsanity.com/paper/2606.19552/) (2026)

## Abstract (arXiv)

Neural language models (LMs) are arguably less data-efficient than humans from a language acquisition perspective. One fundamental question is why this human-LM gap arises. This study explores the advantage of grounded language acquisition, specifically the impact of visual information -- which humans can usually rely on but LMs largely do not have access to during language acquisition -- on syntactic generalization in LMs. Our experiments, following the poverty of stimulus paradigm under two scenarios (using artificial vs. naturalistic images), demonstrate that if the alignments between the linguistic and visual components are clear in the input, access to vision data does help with the syntactic generalization of LMs, but if not, visual input does not help. This highlights the need for additional biases or signals, such as mutual gaze, to enhance cross-modal alignment and enable efficient syntactic generalization in multimodal LMs.