Search papers, labs, and topics across Lattice.
5
1
6
10
Object-centric mask conditioning in MaskWAM dramatically improves policy performance, outperforming traditional WAMs by effectively reducing language ambiguity in complex environments.
Training with just 3% of high-quality motion data can yield superior tracking performance compared to using the entire dataset.
Humanoid-GPT achieves unprecedented zero-shot generalization in motion tracking, outperforming traditional models by leveraging a billion-scale motion dataset.
By explicitly modeling spatial interaction and manipulation intent, AIM achieves a 94% success rate on RoboTwin 2.0, suggesting that robot control can be drastically improved by explicitly reasoning about *where* to interact, not just *how* scenes evolve.
Robots learn better when they first imagine the future and then figure out how to act, unlocking SOTA performance by disentangling forward and inverse dynamics pretraining.
