Search papers, labs, and topics across Lattice.
5
0
8
0
Recursive search can unlock substantial gains in long-video understanding, outperforming traditional hand-crafted approaches.
ZID reveals critical distributional insights that FID and KID overlook, enabling researchers to diagnose generative model failures with unprecedented precision.
Current action-conditioned world models fail to reliably follow diverse off-expert actions, risking the effectiveness of policy learning in real-world applications.
System Intelligence emerges as a game-changer, enabling LLM agents to collaborate effectively across complex tasks by organizing their interactions through dynamic graph structures.
Re-ranking control alone boosts key performance metrics by over 2%, but extending it to fine ranking unlocks even greater gains without sacrificing system stability.
