# The Acquisition of Physical Knowledge in Generative Neural Networks

arXiv: https://arxiv.org/abs/2310.19943 · HTML: https://mlsanity.com/paper/2310.19943/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 19 of its 45 references that are in the map.
- First posted on arXiv 30 Oct 2023.

## Builds on (most cited first, 19 of 19)

- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [Deep Unsupervised Learning using Nonequilibrium Thermodynamics](https://mlsanity.com/paper/1503.03585/) (2015)
- [Generative Adversarial Networks](https://mlsanity.com/paper/2203.00667/) (2022)
- [ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness](https://mlsanity.com/paper/1811.12231/) (2018)
- [Learning Latent Dynamics for Planning from Pixels](https://mlsanity.com/paper/1811.04551/) (2018)
- [Interaction Networks for Learning about Objects, Relations and Physics](https://mlsanity.com/paper/1612.00222/) (2016)
- [Using cognitive psychology to understand GPT-3](https://mlsanity.com/paper/2206.14576/) (2022)
- [IntPhys: A Framework and Benchmark for Visual Intuitive Physics Reasoning](https://mlsanity.com/paper/1803.07616/) (2018)
- [Learning Physical Intuition of Block Towers by Example](https://mlsanity.com/paper/1603.01312/) (2016)
- [On the Transfer of Disentangled Representations in Realistic Settings](https://mlsanity.com/paper/2010.14407/) (2020)
- [Cognitive Psychology for Deep Neural Networks: A Shape Bias Case Study](https://mlsanity.com/paper/1706.08606/) (2017)
- [Clockwork Variational Autoencoders](https://mlsanity.com/paper/2102.09532/) (2021)
- [A Survey on Machine Learning Approaches for Modelling Intuitive Physics](https://mlsanity.com/paper/2202.06481/) (2022)
- [Probing Physics Knowledge Using Tools from Developmental Psychology](https://mlsanity.com/paper/1804.01128/) (2018)
- [A Comparative Evaluation of Approximate Probabilistic Simulation and Deep Neural Networks as Accounts of Human Physical Scene Understanding](https://mlsanity.com/paper/1605.01138/) (2016)
- [The developmental trajectory of object recognition robustness: children are like small adults but unlike big deep neural networks](https://mlsanity.com/paper/2205.10144/) (2022)
- [Emulating Human Developmental Stages with Bayesian Neural Networks](https://mlsanity.com/paper/1902.07579/) (2019)

## Built on by (most cited first, 3 of 3)

- [Visual cognition in multimodal large language models](https://mlsanity.com/paper/2311.16093/) (2023)
- [YoCausal: How Far is Video Generation from World Model? A Causality Perspective](https://mlsanity.com/paper/2605.30346/) (2026)
- [Next state prediction gives rise to entangled, yet compositional representations of objects](https://mlsanity.com/paper/2410.04940/) (2024)

## Abstract (arXiv)

As children grow older, they develop an intuitive understanding of the physical processes around them. Their physical understanding develops in stages, moving along developmental trajectories which have been mapped out extensively in previous empirical research. Here, we investigate how the learning trajectories of deep generative neural networks compare to children's developmental trajectories using physical understanding as a testbed. We outline an approach that allows us to examine two distinct hypotheses of human development - stochastic optimization and complexity increase. We find that while our models are able to accurately predict a number of physical processes, their learning trajectories under both hypotheses do not follow the developmental trajectories of children.