# Meta-learning as a principle for human-like visual representations

arXiv: https://arxiv.org/abs/2606.28399 · HTML: https://mlsanity.com/paper/2606.28399/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 8 of its 64 references that are in the map.
- First posted on arXiv 24 Jun 2026.

## Builds on (most cited first, 8 of 8)

- [Layer Normalization](https://mlsanity.com/paper/1607.06450/) (2016)
- [Sparse Autoencoders Find Highly Interpretable Features in Language Models](https://mlsanity.com/paper/2309.08600/) (2023)
- [DINOv3](https://mlsanity.com/paper/2508.10104/) (2025)
- [SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features](https://mlsanity.com/paper/2502.14786/) (2025)
- [Universal linguistic inductive biases via meta-learning](https://mlsanity.com/paper/2006.16324/) (2020)
- [Prisma: An Open Source Toolkit for Mechanistic Interpretability in Vision and Video](https://mlsanity.com/paper/2504.19475/) (2025)
- [Human-like Category Learning by Injecting Ecological Priors from Large Language Models into Neural Networks](https://mlsanity.com/paper/2402.01821/) (2024)
- [Meta-learning ecological priors from large language models explains human learning and decision making](https://mlsanity.com/paper/2509.00116/) (2025)

## Abstract (arXiv)

The structure of human visual representations underpins our capacity for adaptive behaviour. While pretrained neural networks model human visual representations with unprecedented success, a large discrepancy remains. We propose one reason: these networks optimise a single fixed objective, whereas human representations must support open-ended tasks. We hypothesise this flexibility arises from meta-learning (learning to learn), a pressure shaping representations to acquire new tasks from few observations. To test this, we train a sequence model, without any supervision from human data, across thousands of semantically rich tasks mapping images to high-level concepts. Compared to their pretrained base encoders, meta-learned representations better predict human similarity judgements, semantic rule learning, and high-level visual cortex. Behavioural gains depend on disentangled, high-level task distributions, while brain alignment is driven primarily by the learning-to-learn pressure. Our results suggest the flexibility of human visual representations reflects the functional demand to learn new semantic relationships on the fly.