Search papers, labs, and topics across Lattice.
1
0
3
Disentangling perception and reasoning with role-specific rewards in multimodal LLMs boosts accuracy by 7 points, revealing a critical bottleneck in existing joint optimization approaches.
