Search papers, labs, and topics across Lattice.
ETH Z眉rich
3
0
3
17
MT models show a troubling bias towards US content, with specialized models proving less robust across diverse locales.
Adaptive sampling can cut human evaluation costs while boosting the accuracy of model rankings in NLP tasks.
Evaluators using cESA can achieve more reliable translation quality assessments while cutting annotation time by leveraging shared context across multiple outputs.
