derive results from logits

• cause the logits are more raw than the sigmoided values
• When to use logits over probabilities?
  • The impact of averaging logits over probabilities on ensembles of neural networks: https://ceur-ws.org/Vol-3215/19.pdf
    • “Due to averaging logits instead of averaging probabilities of stochastic or deterministic networks, the degree of confidence on TPs and FPs increased.”