Search papers, labs, and topics across Lattice.
1
0
3
1
LLMs struggle to generate valid and executable ISO-GQL queries, achieving only 4% execution accuracy in zero-shot settings, highlighting a significant gap in their ability to interact with graph databases.
