# Slot-JEPA: An Object-Structured Prediction Interface for Latent World Models

- Source: HuggingFace Daily Papers
- Published: 2026-10-20T00:00:00.000Z
- RoboHot: https://robohot.news/en/items/bfajk6wobtwhix846cyzenp39
- Original: https://huggingface.co/papers/2610.28074

## Summary

Slot-JEPA decomposes JEPA's dense patch representations into stable semantic slots and time-varying spatial masks, providing a reusable object-structured interface. Built on this interface, the Transition Memory Predictor (TMP) stores visual transitions as compact masks, retrieves context using slot statistics, and uses unbalanced optimal transport (UOT) to supervise mask prediction and evaluate candidate plans.