Search papers, labs, and topics across Lattice.
1
3
2
LLM agents can learn to explore novel states and generalize to new tasks with a hybrid on- and off-policy RL framework that leverages memory.
