1 paper
Tobias Groot, Salo Lacunes, Evgenia Ilia
Natural language generation (NLG) tasks are often subject to inherent variability; e.g. predicting the next word given a context has multiple valid responses, evident when asking m…