Search papers, labs, and topics across Lattice.
Affiliation:
1
0
2
LLM-generated tests are less effective on poorly maintained code, revealing a surprising link between code health and token efficiency.
