Search papers, labs, and topics across Lattice.
5
0
9
2
Projector fine-tuning, commonly used for aligning MLLMs, unexpectedly introduces backdoor vulnerabilities with activation mechanisms distinct from those in text-only LLMs.
GaLa's hypergraph representation reveals hidden semantic relationships in multimodal data, leading to a dramatic boost in procedural planning accuracy.
LALMs are shockingly vulnerable to inaudible audio prompts that can make them execute unauthorized actions, even on commercial systems like Mistral AI and Microsoft Azure.
Over 20 teams vied to decode human attention in video, revealing new insights into saliency prediction techniques.
Audio backdoor attacks leave a tell: triggers are surprisingly stable to destructive noise but fragile to meaning-preserving changes.
