Search papers, labs, and topics across Lattice.
1
0
3
2
LLMs can achieve a 7.5x performance boost in web search and extraction by using a bi-level multi-agent architecture with iterative refinement and shared memory.
