Search papers, labs, and topics across Lattice.
1
0
3
Forget expensive human annotations: LLMs can reliably generate synthetic data to validate NLP evaluation metrics, even outperforming human agreement in some multilingual tasks.
