# Modeling The Object Representations Underlying Human Physical Reasoning

arXiv: https://arxiv.org/abs/2602.12486 · HTML: https://mlsanity.com/paper/2602.12486/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 18 of its 64 references that are in the map.
- First posted on arXiv 12 Feb 2026.

## Builds on (most cited first, 18 of 18)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [The Cityscapes Dataset for Semantic Urban Scene Understanding](https://mlsanity.com/paper/1604.01685/) (2016)
- [A ConvNet for the 2020s](https://mlsanity.com/paper/2201.03545/) (2022)
- [Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation](https://mlsanity.com/paper/1802.02611/) (2018)
- [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://mlsanity.com/paper/2105.15203/) (2021)
- [V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation](https://mlsanity.com/paper/1606.04797/) (2016)
- [On the Spectral Bias of Neural Networks](https://mlsanity.com/paper/1806.08734/) (2018)
- [Unified Perceptual Parsing for Scene Understanding](https://mlsanity.com/paper/1807.10221/) (2018)
- [Why Does Surprisal From Larger Transformer-Based Language Models Provide a Poorer Fit to Human Reading Times?](https://mlsanity.com/paper/2212.12131/) (2022)
- [Learning Physical Intuition of Block Towers by Example](https://mlsanity.com/paper/1603.01312/) (2016)
- [Harmonizing the object recognition strategies of deep neural networks with humans](https://mlsanity.com/paper/2211.04533/) (2022)
- [Learning Visual Predictive Models of Physics for Playing Billiards](https://mlsanity.com/paper/1511.07404/) (2015)
- [Transformer-Based Language Model Surprisal Predicts Human Reading Times Best with About Two Billion Training Tokens](https://mlsanity.com/paper/2304.11389/) (2023)
- [Newtonian Image Understanding: Unfolding the Dynamics of Objects in Static Images](https://mlsanity.com/paper/1511.04048/) (2015)
- [Performance-optimized deep neural networks are evolving into worse models of inferotemporal visual cortex](https://mlsanity.com/paper/2306.03779/) (2023)
- [Modeling human intuitions about liquid flow with particle-based simulation](https://mlsanity.com/paper/1809.01524/) (2018)
- [Object permanence in newborn chicks is robust against opposing evidence](https://mlsanity.com/paper/2402.14641/) (2024)

## Abstract (arXiv)

Humans appear to represent objects when reasoning about physics with coarse, volumetric "bodies" that smooth concavities, trading fine visual detail for efficient physical predictions. Yet, the structure of these representations remains largely unknown. Segmentation models, in contrast, are trained for pixel-accurate masks that may misalign with such bodies. We ask whether and when these models nonetheless acquire human-like object representations. Using a time-to-collision (TTC) and change detection (CD) behavioral task with data from 178 and 50 human participants, respectively, we introduce a pipeline and an alignment metric to compare the visual representations of segmentation models to those of humans. We do this systematically on multiple architectures (DINOv2, SegFormer, DeepLabV3+, and UPerNet), varying their size and training time. We find that briefly trained models segment objects too coarsely, aligning poorly with humans, while fully trained models segment objects too finely. For each model, there is an intermediate training regime that best matches the coarse bodies observed in human behaviour, and larger models tend to reach it earlier. We show these bodies emerge under resource constraints in general-purpose vision models, providing computational support to resource-rational accounts of human cognition. This work provides a foundational framework for testing alignment between vision models and humans and shows there is a growing gap between the state-of-the-art in artificial intelligence and human cognition, driven by scaling model size and training.