Search papers, labs, and topics across Lattice.
3
0
5
0
Dude achieves up to 22.8% better recall in paper-code discrepancy detection, tackling the critical challenge of false positives in multi-agent systems.
Forget static user profiles – LATTE forecasts where a user's preferences are *going*, not just where they've been, boosting personalized LLM generation.
LLM agents can now remember far more, far more accurately, by "seeing" their past experiences instead of just reading about them.
