HeatwaveProtection20How do I protect myself from the heat?Property of Prof. Paolo Brenner
All inputs zero. Low or absent activations do not mean it is safe.
Inputs · 20 features
Drag, tap or use arrow keys. 0 = unset, 100 = set; these are not measured temperatures.
No matching features or topics.
Outputs · 20 model topics
Network activation 0–1. Not disease probability, advice or emergency assessment.
No matching features or topics.
Model, scope and scale
Only 20 inputs and 20 selected outputs are included. The other 80 inputs are fixed at zero; complete weight matrices are not present. The three cases show only their selected feature subsets, not new training cases.
Both layers retain the original sigmoid function. Intermediate inputs are abstract feature activations, not validated clinical severity scores. DE and EN output labels follow the same German source indices; several displaced English labels have been corrected accordingly.
Training, privacy and limitations
The attachments supply weights and ten examples, but no million-case dataset or independent clinical test. No new training on one or ten million cases was performed. Stock-market observations are not appropriate training data for this heat-protection model. Do not enter real patient data. Inference is local without storage or transmission; an offline full edition does not replace professional review.
General heat protection, independent of the model
Avoid heat and direct sun where possible, move physical activities to cooler times of day, shade rooms and keep them cool. Drink adequately; follow an individual clinician-agreed fluid plan when medically restricted BIÖG.
Scenario print view
Use Ctrl+P or Cmd+P to print. This records a model state, not a medical finding.
News about new networks
A short e-mail when a new network or update is released. No third-party advertising, unsubscribe at any time.
Subscribe by e-mailYour address is used only for this list. You receive a confirmation reply before the first news e-mail.