Search papers, labs, and topics across Lattice.
These authors contributed equally to this work.Corresponding author
3
0
6
0
LVLMs may excel at describing scenes but often falter in causal reasoning, revealing a critical gap in their understanding of first-person visual safety.
PFL methods are alarmingly vulnerable to adversarial attacks, with malicious clients capable of sabotaging peer models through crafted examples.
Achieving differential privacy in multi-objective submodular maximization opens new avenues for handling sensitive data without sacrificing performance.
