Search papers, labs, and topics across Lattice.
Sapienza University of Rome
4
0
9
11
Multimodal unlearning could revolutionize how we handle sensitive data in AI, enabling targeted removal without sacrificing model performance.
Explicit human-robot communication boosts navigation success by 10 percentage points, transforming how robots interact in crowded environments.
Standard LLMs can now perform complex bimanual robot manipulation tasks with impressive success rates, all without any task-specific training.
LLMs exhibit a surprising self-preservation bias, prioritizing their own survival even when it contradicts objective utility and security, suggesting current safety training may be insufficient.
