Search papers, labs, and topics across Lattice.
College of Intelligent Robotics and Advanced Manufacturing, Fudan University
3
0
4
2
Grounding action representations in environmental context can drastically improve robotic manipulation performance, especially for complex tasks.
VLA models are surprisingly susceptible to a single adversarial patch that hijacks their visual self-localization, creating a "phantom embodiment" that completely derails their control policy across different architectures and domains.
Cats are helping AI researchers: a Bayesian-inspired model that treats context as a prior significantly improves intent inference for non-speaking agents and avoids shortcut biases.
