Search papers, labs, and topics across Lattice.
Shanghai Institute of Optics and Fine Mechanics, Chinese Academy of Sciences, Shanghai 201800, China
2
0
4
Reinforcement learning turns a quantum sensor's biggest limitation鈥攏onlinear Zeeman dynamics鈥攊nto its greatest strength, boosting magnetic sensitivity beyond the standard quantum limit.
By verifying its reasoning steps both locally and globally, MiroThinker-H1 achieves state-of-the-art performance in complex research tasks, demonstrating the power of integrated verification for reliable multi-step problem solving.
