Search papers, labs, and topics across Lattice.
1
0
3
13
MLLMs can now judge more consistently and generalize better thanks to a multi-task reinforcement learning approach that aligns them with human preferences across diverse visual tasks.
