Search papers, labs, and topics across Lattice.
\thepa
2
20
5
13
Forget slow, white-box optimization: VQQA uses a multi-agent system to ask questions about generated videos, using the answers as semantic gradients to drastically improve video quality via black-box prompt refinement.
Forget hand-annotated data: Magnet distills multi-turn tool-use skills into LLMs by automatically generating training trajectories that outperform even Gemini 1.5 Pro.
