Search papers, labs, and topics across Lattice.
9
0
10
6
Progress reward modeling could revolutionize how robots learn in dynamic environments by providing nuanced feedback beyond mere task completion.
RoboDojo reveals that integrating simulation and real-world tasks can significantly enhance the evaluation of robot manipulation policies, bridging the gap between theoretical performance and practical deployment.
Student models can guide the creation of complex teacher models, yielding unexpected computational efficiencies and improved performance.
Current VLMs miss crucial transition-level physics, but APT-Tune teaches them to learn causal transitions without forgetting event-level context.
MagicSim revolutionizes robot learning by merging diverse world construction and execution into a single, efficient framework that enhances both evaluation and interaction capabilities.
Achieving superior annotation efficiency and task success rates, AnnotateAnything revolutionizes how 3D assets are prepared for robot manipulation.
VLMs struggle to meaningfully ground numerical outputs in spatial contexts, often performing at chance levels in critical tasks.
Forget complex assembly: this 3D printing technique lets you pop out functional, self-folding robots with integrated sensors and actuators directly from a flat sheet.
Finally, a robot can reliably pick out that specific shirt from a messy pile, thanks to a new vision-language pipeline that reasons about garment affordances.
