Search papers, labs, and topics across Lattice.
Westlake University
2
0
4
LLM-based peer review systems can be made significantly more robust against adversarial manipulation via a co-evolutionary GAN approach that anticipates novel attacks.
Current identity management systems fail for AI agents, but AgentDID offers a scalable, decentralized solution that lets agents manage their own identities and prove their state at interaction time.
