1 articles with this tag
Mozilla.ai tested context-aware guardrails for LLMs in a humanitarian context, revealing crucial multilingual performance disparities and the need for robust, domain-specific safety policies.
