Browse personalized knowledge signals. Useful reads and comments become recommendation inputs for future people matches.
Safety note: XinFora AI is a professional knowledge-sharing and collaboration platform. It does not provide therapy, diagnosis, medical advice, treatment, clinical consultation, crisis counseling, or emergency services.
OpenAI described ongoing mental-health-related safety work, including well-being governance, family protections, and a planned trusted contact feature.
Useful for tracking how large AI platforms are framing mental-health safety boundaries.
Sword Health introduced MindGuard, a family of safety classifiers and evaluation assets for mental-health AI conversations, with explicit research and human-oversight limitations.
A close public example of safety infrastructure rather than a therapy product.
A 2024 open-access systematic review covering applications of large language models in psychiatry and highlighting limitations across study design, data, samples, and evaluation.
A strong seed paper for researchers and builders entering Mental Health x AI.
Public code and tutorial materials accompanying a medRxiv primer on benchmarking language models for clinical safety with mental-health professionals.
Directly useful for builders who want reproducible safety evaluation rather than marketing claims.
An open-source evaluation pipeline that simulates multi-turn conversations and scores model behavior on six clinical safety criteria.
A close public example of the kind of safety-evaluation work XinFora should surface.
A 2025 Frontiers paper evaluating 15 LLMs on mental-health knowledge and illness-diagnosis tasks in a Chinese context, using public datasets and qualification-exam questions.
Important for XinFora's China-US positioning because it focuses on Chinese mental-health scenarios.