Aymara AI evaluates 20 commercial LLMs across 10 safety domains via policy-grounded adversarial prompts, reporting mean safety scores from 86.2% to 52.4% with large domain-specific gaps.
A coefficient of agreement for nominal scales
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
representative citing papers
citing papers explorer
-
Policy-Grounded Safety Evaluation of 20 Large Language Models
Aymara AI evaluates 20 commercial LLMs across 10 safety domains via policy-grounded adversarial prompts, reporting mean safety scores from 86.2% to 52.4% with large domain-specific gaps.
- Hilbert-Geo: Solving Solid Geometric Problems by Neural-Symbolic Reasoning