Search papers, labs, and topics across Lattice.
The Chinese University of Hong Kong
1
0
3
1
Compress recommender system embedding tables by 75% with minimal accuracy loss using a balanced co-clustering approach that's also orders of magnitude faster than alternatives.
