Search papers, labs, and topics across Lattice.
1
0
2
Forget hand-engineering reward machines: this work learns them directly from raw state trajectories, even without reward or label information.
