# Grid or Nodes? Probing Object Memory in World Models Under Partial Observability

arXiv: https://arxiv.org/abs/2610.18637 · HTML: https://mlsanity.com/paper/2610.18637/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Latent predictive world models, one of 11 directions in World models.
- Builds on 30 of its 32 references that are in the map.
- First posted on arXiv 13 Oct 2026.

## Placed in

- World models / [Latent predictive world models](https://mlsanity.com/wm/latent-predictive-world-models/)

## Builds on (most cited first, 30 of 30)

- [Object-Centric Learning with Slot Attention](https://mlsanity.com/paper/2006.15055/) (2020)
- [Probing Classifiers: Promises, Shortcomings, and Advances](https://mlsanity.com/paper/2102.12452/) (2021)
- [Simple Online and Realtime Tracking](https://mlsanity.com/paper/1602.00763/) (2016)
- [V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning](https://mlsanity.com/paper/2506.09985/) (2025)
- [Performance Measures and a Data Set for Multi-Target, Multi-Camera Tracking](https://mlsanity.com/paper/1609.01775/) (2016)
- [Designing and Interpreting Probes with Control Tasks](https://mlsanity.com/paper/1909.03368/) (2019)
- [Emergent World Representations: Exploring a Sequence Model Trained on a Synthetic Task](https://mlsanity.com/paper/2210.13382/) (2022)
- [DINO-WM: World Models on Pre-trained Visual Features enable Zero-shot Planning](https://mlsanity.com/paper/2411.04983/) (2024)
- [Minigrid & Miniworld: Modular & Customizable Reinforcement Learning Environments for Goal-Oriented Tasks](https://mlsanity.com/paper/2306.13831/) (2023)
- [Contrastive Learning of Structured World Models](https://mlsanity.com/paper/1911.12247/) (2019)
- [WorldMem: Long-term Consistent World Simulation with Memory](https://mlsanity.com/paper/2504.12369/) (2025)
- [Learning to Track with Object Permanence](https://mlsanity.com/paper/2103.14258/) (2021)
- [Evaluating the World Model Implicit in a Generative Model](https://mlsanity.com/paper/2406.03689/) (2024)
- [Shaping Belief States with Generative Environment Models for RL](https://mlsanity.com/paper/1906.09237/) (2019)
- [Memory Forcing: Spatio-Temporal Memory for Consistent Scene Generation on Minecraft](https://mlsanity.com/paper/2510.03198/) (2025)
- [World-in-World: World Models in a Closed-Loop World](https://mlsanity.com/paper/2510.18135/) (2025)
- [Emergence of Maps in the Memories of Blind Navigation Agents](https://mlsanity.com/paper/2301.13261/) (2023)
- [Learning Object Permanence from Video](https://mlsanity.com/paper/2003.10469/) (2020)
- [LiveWorld: Simulating Out-of-Sight Dynamics in Generative Video World Models](https://mlsanity.com/paper/2603.07145/) (2026)
- [Out of Sight but Not Out of Mind: Hybrid Memory for Dynamic Video World Models](https://mlsanity.com/paper/2603.25716/) (2026)
- [Unsupervised Object-Based Transition Models for 3D Partially Observable Environments](https://mlsanity.com/paper/2103.04693/) (2021)
- [Latent Particle World Models: Self-supervised Object-centric Stochastic Dynamics Modeling](https://mlsanity.com/paper/2603.04553/) (2026)
- [Out of Sight, Out of Mind? Evaluating State Evolution in Video World Models](https://mlsanity.com/paper/2603.13215/) (2026)
- [Current World Models Lack a Persistent State Core](https://mlsanity.com/paper/2606.20545/) (2026)
- [MemoBench: Benchmarking World Modeling in Dynamically Changing Environments](https://mlsanity.com/paper/2606.27537/) (2026)
- [LoopNav: Benchmarking Spatial Consistency in World Models](https://mlsanity.com/paper/2505.22976/) (2025)
- [Flow Equivariant World Models: Memory for Partially Observed Dynamic Environments](https://mlsanity.com/paper/2601.01075/) (2026)
- [Echo-Memory: A Controlled Study of Memory in Action World Models](https://mlsanity.com/paper/2606.09803/) (2026)
- [Object-Centric World Models for Causality-Aware Reinforcement Learning](https://mlsanity.com/paper/2511.14262/) (2025)
- [R2M-Bench: Evaluating Revisit Memory via Relative Consistency in Interactive Video World Models](https://mlsanity.com/paper/2608.27328/) (2026)

## Abstract (arXiv)

A robot's world model must remember objects that leave the camera's view. FloWM, a learned world model, keeps them in a spatial grid of features that it moves with five built-in velocities. A node memory instead keeps each object's position and velocity as numbers. We compare them in FloWM's own block world, with an oracle-fed Kalman tracker as the node memory. How the grid is read changes its score: a single-cell probe finds 66% of out-of-view blocks, while FloWM's own probe 89%. FloWM also scores 8 points higher in view although no frame is shown, which we attribute to its write-back of its own predictions. Out of view, the tracker reaches FloWM's range in 14KB against 5.6MB, with no probe. On diagonal motion, outside its five velocities, FloWM finds 11% of blocks even after fine-tuning, while the unchanged tracker keeps 93%. Faster blocks defeat both until one more allowed speed takes the tracker to 94%, where FloWM's built-in shifts would need 25 copies of its grid. A learned node memory must still replace the oracles and handle velocity under noise, long gaps, and the bounce.