Unverified50% confidenceFactExact time
HealthBench is a benchmark oriented toward real-world medical dialogue scenarios, where questions are mostly open-ended consultations and answers have no single correct form.
1
Sources
50%
Confidence
Long-term
Relevance
10/30/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedMentalHealthBench covers realistic mental health conversations, including ambiguous expressions, multi-turn interactions, and emotional fluctuations.73% similarUnverifiedHealthBench Professional是专门针对临床医疗场景的评估体系，要求模型在专业知识准确性、伦理边界和不确定性表达上达标69% similarUnverifiedMentalHealthBench is an evaluation benchmark designed with expert involvement to measure whether AI systems' responses in real mental health conversation scenarios are both helpful and sufficiently safe.68% similarUnverifiedKin Health 不提供逐字逐句的转录文本，而是将对话内容重新组织成聚焦的结构化摘要，整理出关键要点、医生建议和后续行动步骤62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1114370API
curl https://kongchang.com/api/v1/knowledge/claims/1114370MCP
get_claim(id=1114370)