Search papers, labs, and topics across Lattice.
1
0
3
2
LLMs can be taught to "think longer" and explore more diverse reasoning paths in-context via a simple length-incentivized reward, leading to improved generalization.
