Search papers, labs, and topics across Lattice.
Shenzhen Campus of Sun Yat-sen University, Shenzhen Loop Area Institute
3
0
4
A single adaptive framework can boost the efficiency of zeroth-order optimization by up to 3x without increasing memory usage.
Guaranteeing injectivity in autoencoders unlocks more robust and geometrically faithful dimensionality reduction, even under distribution shift.
DNNs can be both more accurate *and* better calibrated: Generative Cross-Entropy (GCE) offers a way to escape the accuracy-calibration trade-off.
