Search papers, labs, and topics across Lattice.
1
0
3
2
Current RAG evaluation metrics fail to capture the nuances of multi-hop reasoning, but a new context-aware metric closes the gap.
