Search papers, labs, and topics across Lattice.
Peking University
1
0
2
8
Forget fine-tuning: large pretrained models are already teeming with task-specific experts densely packed around their weights, just waiting to be discovered via simple random sampling.
