Search papers, labs, and topics across Lattice.
1
0
4
3
LLM judges in human-AI coding collaborations show surprisingly low inter-rater reliability, suggesting current evaluation methods may be inadequate for assessing true co-creation effectiveness.
