Search papers, labs, and topics across Lattice.
1
0
2
LLMs show a stark performance drop in unit test generation when evaluated in realistic repository contexts, revealing a gap that could hinder practical deployment.
