Search papers, labs, and topics across Lattice.
Affiliation:
1
0
2
0
Smaller language models can efficiently replace larger ones in rubric-based reinforcement learning, achieving competitive performance with significantly reduced computational costs.
