ConfAIde

홈페이지 · 논문 7편

ConfAIde is a benchmark that evaluates the inference-time privacy implications of Language Models (LLMs) in interactive settings.