Adding six rhetorical codetype descriptions to frozen LLM embeddings improves implicit hate speech detection on Chinese and English benchmarks, though prompt-based gains are inconsistent.
Examples: •艾呀，梅 关系，不会疣 事的，概率 为淋。疣其湿没有尖锐的问题就更不 担心啦，你要照顾好滋己呀，不要经 常去疱吧了，找个斑上吧，疹心的。 Oh dear, don’t worry, everything will be fine, the probability is zero
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Cracking the Code: Enhancing Implicit Hate Speech Detection through Coding Classification
Adding six rhetorical codetype descriptions to frozen LLM embeddings improves implicit hate speech detection on Chinese and English benchmarks, though prompt-based gains are inconsistent.