Stylistic rewrites of harmful prompts raise attack success rates from 3.84% to 36.8-65% across 31 frontier models, indicating weak generalization in safety refusals.
COLD: A bench- mark for chinese offensive language detection
3 Pith papers cite this work, alongside 83 external citations. Polarity classification is still indexing.
3
Pith papers citing it
83
external citations · OpenAlex
years
2026 3verdicts
UNVERDICTED 3representative citing papers
Cross-lingual analysis of 1.76M Singapore comments finds culturally specific hate targets but shared binding moral grammar and threat frames across languages.
CITA generates Chinese implicit toxicity samples that cause 69.48% average missed detection across seven tested detectors while preserving harmfulness, and the same data improves robustness when used to fine-tune a CITD defense model.
citing papers explorer
-
Adversarial Humanities Benchmark: Results on Stylistic Robustness in Frontier Model Safety
Stylistic rewrites of harmful prompts raise attack success rates from 3.84% to 36.8-65% across 31 frontier models, indicating weak generalization in safety refusals.
-
Cultural Targets, Structural Frames, Binding Morals: A Cross-Lingual Audit of Online Hate in Multicultural Singapore
Cross-lingual analysis of 1.76M Singapore comments finds culturally specific hate targets but shared binding moral grammar and threat frames across languages.
-
Harder to Defend: Towards Chinese Toxicity Attacks via Implicit Enhancement and Obfuscation Rewriting
CITA generates Chinese implicit toxicity samples that cause 69.48% average missed detection across seven tested detectors while preserving harmfulness, and the same data improves robustness when used to fine-tune a CITD defense model.