Search papers, labs, and topics across Lattice.
Beijing University of Posts and Telecommunications
3
0
7
LookAgain's innovative closed-loop approach allows for real-time refinement of GUI grounding predictions, drastically improving accuracy on difficult tasks.
Offline model-based optimization is fundamentally a ranking problem, and focusing on ranking near-optimal designs beats traditional regression-based surrogate modeling.
By intelligently weighting prototypes based on their uncertainty, CAFedCL sidesteps the "prototype bias loop" that plagues federated learning with imbalanced data, leading to significant gains in both accuracy and fairness.
