// multi-utility computation suite · offline · instant · precise
┌──────────────────────────┐
│ [c] calcalyst_ │
│ computation suite │
└──────────────────────────┘
// select a module to initialize
/ search↵ open firstesc close
// adsenseEMPTY_LEADER_SLOT728×90
// adsenseMOBILE_ANCHOR_SLOT320×50
// keyboard_shortcuts
/focus search
↑↓navigate module list
Enter
open first result from search
open highlighted
compute when module is open
compute when focused in a field
Escclose module · clear selection
⌫
sci.softmax-neural-network Calculator
Calculates softmax output probabilities from neural network logits: σ(z)_i = exp(z_i)/Σexp(z_j), and cross-entropy loss for classification training. Softmax with temperature T: dividing logits by T > 1 produces flatter distributions (knowledge distillation); T < 1 sharpens predictions — used in LLM token sampling.
Inputs
Z1
Reference formula or conversion factor shown for context.
Z2
Reference formula or conversion factor shown for context.
Z3
Reference formula or conversion factor shown for context.
Temperature
Thermal state of the substance. Check whether the formula needs Celsius, Fahrenheit, or Kelvin (K = °C + 273.15).
Results
softmax p₁ (class 1)
The classification assigned based on the computed value.
softmax p₂ (class 2)
The classification assigned based on the computed value.
softmax p₃ (class 3)
The classification assigned based on the computed value.
output entropy H (bits)
Sample size or count used in the calculation.
σ(z)_i = exp(z_i/T) / Σexp(z_j/T)
Reference formula or conversion factor shown for context.
temperature effect
The change in the property attributable to the temperature difference. Many material properties are temperature-dependent.