Search papers, labs, and topics across Lattice.
3
0
8
0
RepMT-SAC achieves up to 30% better performance in multi-task reinforcement learning by leveraging a task-agnostic core for efficient skill transfer.
A new measurement-manifold compatibility measure reveals how to improve generative model trustworthiness in medical imaging by guiding more effective sampling strategies.
Frontier models can't build playable games in one shot, but a closed-loop system using GUI agents to playtest and refine code achieves a 66.8% success rate, proving that game generation needs to be a conversation, not a translation.
