hidden layer
A layer of neurons between the input and output of a neural network that builds internal, intermediate representations of the input data.
As defined in 68 | Melanie Mitchell on Artificial Intelligence and the Challenge of Common Sense
Where it comes up · 2
- Robert Burton on Being Certain5 claims
- 68 | Melanie Mitchell on Artificial Intelligence and the Challenge of Common Sense3 claims
Most important claims · 8 of 8Search all claims for “hidden layer” →
- Two people witnessing the same event with identical sensory inputs will perceive it differently, because retinal-level input is reshaped through hierarchical layers of the visual cortex and neural networks shaped by genetics, biological predispositions (fear, optimism, pessimism), prior experience, and even epigenetic inheritance—so expecting two brains to produce the same output from the same input is a fundamental fallacy of rationality.Robert Burton on Being Certain
- Backpropagation, developed in the late 1970s and based on calculus, solved the previously unsolved problem of how to apportion credit and blame for output errors back through hidden layers, enabling the training of multi-layer neural networks and disproving Minsky and Papert's doubt that such an algorithm could exist.68 | Melanie Mitchell on Artificial Intelligence and the Challenge of Common Sense
- Neural networks gained power over perceptrons because hidden layers can construct intermediate internal representations of input — like the concept of a loop in a handwritten digit — capturing global holistic properties rather than single-pixel features, roughly analogous to what the visual cortex does.68 | Melanie Mitchell on Artificial Intelligence and the Challenge of Common Sense
- Even though you cannot fully fix the brain using the brain, you can become more aware and sensitive to how your priors, experience, and predilections shape your conscious world—as Roberts did by later doubting the Challenger study he had eagerly believed—without ever fully understanding the underlying processes.Robert Burton on Being Certain
- Thoughts and decisions emerge from thousands of embedded traits in the neural network each 'voting' on an output; the brain separately calculates a likelihood of correctness, and only when that calculated feeling of knowing exceeds a threshold does the idea reach consciousness—explaining why the name 'Pogo' surfaces fully formed after sleep without conscious effort.Robert Burton on Being Certain
- If people recognized that they operate under principles they don't understand and hidden layers firing in ways they can't control—and learned roughly how prone they are to certainty, risk aversion, or optimism—they could accept that others with different tendencies will see the world entirely differently, which would defuse much unnecessary division like the conflict over Trump.Robert Burton on Being Certain
- Things go on beneath conscious awareness that we have no access to, as when you go to bed unable to recall a comic strip's name and wake with 'Pogo' in mind, or drive a familiar route while daydreaming—your brain works on tasks and executes them without conscious will.Robert Burton on Being Certain
- A deep neural network is simply a neural network with more than two hidden layers; the word 'deep' refers to architectural depth, not profundity of thought.68 | Melanie Mitchell on Artificial Intelligence and the Challenge of Common Sense