Search papers, labs, and topics across Lattice.
1
0
3
1
Smaller reasoning models can achieve both higher accuracy and shorter reasoning chains by adaptively penalizing unnecessary reflections and coordinating length penalties with problem complexity.
