Search papers, labs, and topics across Lattice.
1
0
2
Multi-agent debate unlocks significantly better argument classification from LLMs, even without fine-tuning, by surfacing hidden ambiguities that single-agent models miss.
