Search papers, labs, and topics across Lattice.
1
0
2
A revelation principle for AI agents reveals how to incentivize honesty and obedience even when their true capabilities are hidden.
