1 paper
Stefan F. Schouten, Ilia Markov, Piek Vossen
We introduce a new, contextual, multilingual dataset called ToxiREX: Toxic REasoning in ConteXt. The dataset consists of threads of Reddit comments and structured characterizations…