Search papers, labs, and topics across Lattice.
1
0
3
1
Ditching the always-on LLM for proactive agents can boost speed by up to 83x and improve F1 scores by 16.7, simply by processing structured event data directly with a temporal graph learning model.
