Search papers, labs, and topics across Lattice.
Vrije Universiteit Amsterdam
1
0
3
2
Forget expensive LLM inference for MTQE: train a COMET model on GPT-4o-generated annotations and get competitive performance.