Embedding-based clustering of five safety benchmarks reveals six rough harm themes, with datasets showing uneven topic coverage such as GretelAI on privacy and WildGuardMix on self-harm.
(2021) ”On the dangers of stochastic parrots: Can languagemodels be too big?”, Proceedings of the 2021 ACM conference on fairness, accountability, and transparency
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Surfacing Semantic Orthogonality Across Model Safety Benchmarks: A Multi-Dimensional Analysis
Embedding-based clustering of five safety benchmarks reveals six rough harm themes, with datasets showing uneven topic coverage such as GretelAI on privacy and WildGuardMix on self-harm.