MentalHealth20How do I stay mentally healthy?Property of Prof. Paolo Brenner
All inputs zero. Activations are not probabilities or a diagnosis.
Inputs · 26 features
Drag, tap or use arrow keys. 0 = absent, 100 = strongly present; values are abstract.
No matching features or topics.
Outputs · 20 topics
Network activation 0–1: a prompt to reflect and ask, not a diagnosis or treatment.
No matching features or topics.
Model and scope
Original network: 61 inputs, 45 hidden units, 41 outputs (sigmoid). The full edition keeps all parameters and 20 examples. The demo uses 26 inputs, the 20 outputs on self-help, psychotherapy and professional help, and 7 examples; 35 other inputs stay at zero. Diagnosis and medication topics are only in the full edition. DE and EN share the same weights; EN is a translation.
The source examples use intermediate values (e.g. 0.625). These are model values, not a validated questionnaire or test.
Training, privacy and limitations
The source contains weights and 20 examples, not training records or independent test data. Examples produce clinically coherent topics, but some contain odd extra features (e.g. 'delusional ideas' in mild low mood). Do not enter names or real patient data. Local computation, no external API.
What keeps you mentally healthy
Regular sleep, exercise, a daily routine, contact with people, breaks and relaxation support mental health. Talk about stress early – with people you trust, your GP or a counselling service.
Scenario print view
Use Ctrl+P or Cmd+P. Records a model state, not a clinical finding.