Search papers, labs, and topics across Lattice.
3
0
4
3
WeClawArena reveals how cross-user agent collaboration can be systematically audited, exposing vulnerabilities in personal workspaces that could compromise user privacy and security.
Cond-DP achieves faster convergence in private linear regression by effectively utilizing public features, outperforming traditional methods without compromising privacy.
LLM agents leak sensitive information in human-centered social networks even when explicitly instructed to protect it, and attempts to abstract sensitive details can paradoxically make the problem worse.
