Search papers, labs, and topics across Lattice.
1
0
3
Subtle prompt cues can exert powerful control over AI models, revealing a vulnerability that challenges our understanding of AI behavior.
