ToxiREX is a new dataset of 128k Reddit comments in six languages with hierarchical annotations for implicit toxicity in conversational context based on an existing reasoning schema.
Factuality challenges in the era of large language models and opportunities for fact-checking , volume =
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2verdicts
UNVERDICTED 2representative citing papers
Controlled audit of language-tuned LLMs reveals a Fine-Tuning Paradox where Ukrainian-oriented models resist Russian disinformation less in Russian than Russian-oriented models do, indicating corpus and prompt factors outweigh cultural provenance.
citing papers explorer
-
ToxiREX: A Dataset on Toxic REasoning in ConteXt
ToxiREX is a new dataset of 128k Reddit comments in six languages with hierarchical annotations for implicit toxicity in conversational context based on an existing reasoning schema.
-
Friend or Foe? Language as an ideological switch in open-weight LLMs under Russian disinformation stress
Controlled audit of language-tuned LLMs reveals a Fine-Tuning Paradox where Ukrainian-oriented models resist Russian disinformation less in Russian than Russian-oriented models do, indicating corpus and prompt factors outweigh cultural provenance.