Search papers, labs, and topics across Lattice.
1
0
3
LLMs struggle to navigate the nuances of real-world rules, achieving only ~45% accuracy on a new benchmark of legal and policy reasoning tasks.
