Search papers, labs, and topics across Lattice.
1
0
2
6
Adversarial prompt injections can trick LLMs into misclassifying malicious log entries as benign, jeopardizing cybersecurity efforts.
