Search papers, labs, and topics across Lattice.
Peking University
1
0
2
LLMs can't reliably generate the very skills that boost their performance, and smaller models equipped with expert-crafted skills can rival larger, skill-less models.
