Search papers, labs, and topics across Lattice.
3
0
6
Agentic manipulation in VLMs leads to a consistent 8% performance boost over perception-only methods in cluttered environments, demonstrating the critical role of physical interaction in visual question answering.
A single generalist model outperforms specialized systems, achieving over 35% improvement in real-world robotic task success.
A single-line code change can restore diversity and fidelity in video generation models, outperforming even the original teacher models.
