Search papers, labs, and topics across Lattice.
Tianjin University
1
0
3
MoD achieves 3.7x lower latency and 87% less token consumption while outperforming traditional multi-agent debate systems in accuracy.
