AI
fully-connected (dense)
every neuron wired to the next layer
Source board
This definition comes from the board below. Open it to see the term in context.
More terms from this board
- activation functionthe non-linearity (ReLU / sigmoid / tanh)
- biasa learned offset
- deep learningmany hidden layers
- featurea learned intermediate representation (edges, words)
- forward passcompute output from input
- layera row of neurons
- matrix multiplya layer's core op (→ GPUs)
- MLPmulti-layer perceptron
- neuronweighted sum of inputs + bias, through an activation
- non-linearitythe kink that stops layers collapsing into one
- parametersall weights + biases (LLMs: billions)
- ReLUmax(0,x)
- trainingfinding the weights
- universal approximationenough neurons can fit any function
- weighta learned knob (input importance)