Search papers, labs, and topics across Lattice.
1
0
2
Explicitly characterizing adversaries could transform how we evaluate risks associated with releasing open-weight AI models.
