Search papers, labs, and topics across Lattice.
Peking University
1
0
3
Decoupling execution from workflow generation in LLMs yields significantly more accurate and robust structured workflows.