Search papers, labs, and topics across Lattice.
1
0
LLM-generated rubrics can nearly match human evaluators in assessing paper reproduction, but they often miss the mark with excessive granularity and bias.
