Our paper, FORLA: Federated Object-centric Representation Learning with Slot Attention, was accepted to NeurIPS 2025.