Synonyms that are in the dictionary are marked in green. Synonyms that are not in the dictionary are marked in red.
Antonyms that are in the dictionary are marked in green. Antonyms that are not in the dictionary are marked in red.
In addition, we expect to help by minimizing the risk that models generate unsafe or biased results, and we expect to help by hyper-training generalized models for specialized domains.